Gem Tower Defense

Comments

Name:

type hello

216 - 2020-04-06, 10:41 AM

@lement: why you don't make any BO in your GO emerald runs?

I ussually build 1st before 96 if i have GO, because i afraid i got a unlucky stack.
With GO and BO i ussually leak 1 max, sometimes 2 , rarely 3.
Even in my last run which had horrible EE (and even EG) gain, with 5 EE, i leaked 1 in 96 (3 in 100, 1 in 104), maybe 2 GT helped a bit.

i failed this run in 196, because 1+17+0 emeralds,i did even 210 repicks!! (that 1+ mean EE before wawe 20, so only 17 in 387 tries with max chances,it's very unlucky, around 1% chance).
I had over 10% chance to succes in 196 (without few gems, i going to check again when finish my simulator), 18 EE seems just too low.

I see no reason to wait with 1st BO until 190. just build first around 96-110, 2nd arond 140-190, sometimes 3rd around 190. I fell that 2 nd is not very much helpfull, but 1 st reduce amount of leaks by art least 1-2.

I think that lement strat(locations of EE) works well for reducing chance to stack(better than mine).
Chance to (perfect) stack in 96-104 while leaking 2 in very low (around 1% i think), but if leaking more and in later wawes it gets higher. 
I rather was always focusing on trying limiting leaks to 1 (when leaking 1 there is 0% chance to stack :), but withot GO i can't do this.

Being able to survive longer without GO increase chance of getting GO in run, so more chance to 196+ run. 
I'm really suprised that lement survived to 171 without GO AND BO,then got GO.
Recently i got 1st GO in 127 i trough it was late GO...(run fail in 196, 22EE only + a bit unluck+ maybe a bit no optimal EE locations).

Lement's emerals runs looks quite good, definitelly he is able to defeat 196.

If someone (except me) submit legit 196+ score i think i will upload more pics of my runs, i hope more people will be able to do 196+ runs then.

mom - 2020-04-03, 10:15 PM

216 - 2020-04-02, 07:29 PM

216 - 2020-04-02, 07:05 PM

I used your gems in my wawe 196 simulator(i didn't programed some gems jet, so only use EE,EG,jades,RC,and opals).

I distributed about 120 lvls between EEs and EGs, based on shape of maze and my experience(idk what lvl your towers was extractly)

They was leaking 2 ussually, once got 1, sometimes 3-4.

Then added  BO in mid - run 10 times,3x 0 leak, often 2

Then added 3 EE- one in downright place, 2 above your rightmost. - about 50% succes rate.

Then moved one emerald of 2 in righ side (that i added) to place above your malachite - run 10 times, succes 7, 3x 1 leak(once with very low hp, with othere tower it probably even wouldn't happen).

My simulator is still unfinished(because i'm lazy), maybe i need to correct path be few pixels(and add rest important gems), but it work quite well.

So your run definitelly had big chance to be 196+ , sadly get bugged :(

btw, that 5 BS didn't caused big lags? I would uprage last 2.

My times at wawe 20 are around 1000 (slighty above).
at 40 over 2000, maybe around 3000, unless i get BS early , then 20-40 goes fast
at 96 around 10-11k.

Why you didn't submit 181 score? another bug, like mine in 514 run?
I submit all 140+ lvl scores.

216 - 2020-04-02, 05:15 PM

Wow! lvl 182 with no BO?? and you survived to 171 without GO and BO? 
Slighty diffrrent strat than me, but this could be 196+ run, shame that strange bug killed you :( ,i have never had bug like that!

100000 gold left, you should repick a bit more; 12 jades at 135 was enough, building 2 next wasn't needed (unless 1-turn).

looks like it was needed to add few EE in righ part of map, map is too empty around c4, i would also add EE in down right part.

Your maze looks nice, but can be even shorther.
Build 1st BO around 100, it speedup game a bit.
Also that down part of maze is one of reason that slow down ground wawes, ofc that 3 EE on line are well located, but enemies should enter maze around that SYS.

If you removed 1 gem near to your Flawless Diamond (and maybe have BO) ground units wouldn't leak, i think? that righ part of maze may slow down groud wawes(unless you do it intentionally, for more gold)

One of 2 lasts yuor EE is not well positioned(that near bs),That 6 EE between c1 and c5 maybe be too much (5 wans't enough?), i would put something on down line or right part of map.

How many flyers you leaked? (before and after GO)

And how many repicks used?

Anyway u got probably farhtest in emerald run without GO , my record is die in 152.
and you did it even without BO... wow.

You definitelly beat 196 if you don't give up.
That run had everthing needed to be 196+, maybe few more EE was needed, but with your gold that wouldn't be a problem

I never minimalize my game, just use alt + tab.

I thinking what could cause that strange bug? did you have many games/other flash using things open at the same time?
I remember i had a some games(maybe in sanbox) open a few weeks(i hibernate my PC) and eventually some of them got bugged that i couldn't place gems, when clicking "place gem button" i could spam rocks on this button, strage bug.

lement - 2020-04-02, 12:47 PM

Well had an emerald GO run going....On track to 26 at 196, then this happens: https://i.imgur.com/vMjYMmi.png

(Here's two earlier shots at 135 and 175, GO was at 172: https://i.imgur.com/9nSPLEr.png https://i.imgur.com/CG7obS0.png )

Yeah, that was lost; they got bunched up weirdly too...I have another GO emerald run ( https://i.imgur.com/zkiOMkp.png ) with slightly misplaced GO (went to fill a hole, oh well), but not sure that'll succeed; I had to close down other stuff because otherwise the autoclicker nearly freezes game after reducing delay between clicks. 196 might just not be possible for me due that, no matter the strat.

Also had chrome decide to close down and reask me if I want to run flash on another run, that was >_> total of 9 runs with this maze so far I guess, maybe I'll draw something new next time again

@216 yeah, you get far better thimes than me. This run had iirc 2360 20/15095 95/28550 145. Have also gotten over 20k  95 times. It seems time goes faster when game is not minimized, but then lag can mean not getting kills (I once had 50 kills total on Malachite after wave 8...with other gems having like 7 kills, tops.)

For pics, I use imgur (with imgur-screenshot.sh), also have custom bash thing to save copy locally and then use that so I can rename the localy copy with the url I got from it.

GodSama - 2020-03-30, 04:18 AM

216 - 2020-03-26, 05:49 PM

I have few pics, i did sceen while was in 196 first time(while killing last enemy), in 196 with 21 EE and in last run in wawe 500, but didn't upload them (yet), it would be a huge spoiler?

Actually, i have never uploaded a pic, but can i upload somewhere without creating an account?

Also i don't want that everybody will kopy my maze when i upload it, because there are many good mazes that allow to beat 196

i slighty changed maze since since first 196+ run, actually shape of maze isn't much important, only need to keep free space in important locations for EE.

Even amount of EE isn't as important as their location.
If you spam 40 EE in random positions, you lose probably.

My new maze is VERY short, i leak single ground units sometimes if bad run. Good idea is shortening maze during run.

I didn't played since end of 514 run(still keep that game open :), but i plan to get past 500 once again to sumbit score. My time was over 80k, since 250 game was boring, so i will do that only once, then will end 196+ runs intentionally somewhere around 210-300.

btw, firts simulations (actually simulator is not completted, but most most imortant towers are done) shows that defeating 196 without GO is possible, with 32-34 EE, and a bit luck in 196(the same towers can leak 2 and leak 0). 
However getting than many EE, no lose due to stack and having a bit luck in 196 seems much more unlikely than just getting GO.

_mmkay - 2020-03-26, 04:15 PM

216 - 2020-03-23, 10:29 AM

216 - 2020-03-22, 10:18 AM

I realized the game can't last forever - eventually all map will be full, making starting new wawe impossible. But that problems will be around wawe 800-1000, getting that far will last too long and is boring.

Wawe 409 now, time: 62k
I think i will reach max integer value in score(in wawe over 500), then die (i hope game won't crash after exceeding 2.14b score).

Losing in air wawe is very unlikely after 400(even 250-300), with so many stunners(maybe if use opal bug and slow many gems?), i think increasing number of ememies in wawe would keep game being interesting and challenging. Maybe 1 more every 50 wawes. 
12 flyers maybe would be enough to kill me now.

i got 17 Great gems total, kept 11,
including 5 GS (kept 3), and 3 GO. Fist game that i got 5 the same Great gems (even inluding sandbox games). 
Repicks costs over 10k.

If anyone ever will try to submit really highest possible lvl(it take so many DAYS, i think nobody will do this), probably will use "paraiba strat" - it take 4 gems, so map will fill slower(also 1-turn combinations and Great gems wouldn't worth to take).
Note that you can't fill all space except your maze - need to be able to lose. I remove few rocks and block main road in mid of map to allow ground unit go around. I can't see other way to lose(maybe except using opal bug).

What is theorically highest lvl possible? - map is 20x20 ,but checkpoints and maze take some space, so around 300-350 gems is possible to take. with only paraibas it would be lvl 1200-1400, but other gems are also needed, so i think lvl 1000-1200.

Anyway i don't want to do this.
defeating 196 was challenging and give some fun, now i thinking about possibility doing this without GO. I wil simulate if this is possible( i'm writting "wawe 196 sim" in JS).

But game after 250+ is rather boring - after reaching stable point. Maybe always is a small chance to lose, when one lowest priority flyer get far while while all stuner attack others. but since many wawes nothing even readed c6.

i made 6 paraibas to speed up ground wawes - they take around 200s, unless 1 unit get to c4, where is low-dmg stun zone(then around 300s).

Also MASSIVE lag in groung wawes is frustrating - idk why wawes starts ok, then while is less enemies lag more.
Also i remember i had lag when i was sumbitting my 404 run - lad was when click name field, and text typed was removed few times! lucklily i didn't click "submit" too fast.
I think i will wait few minutes before tring to click name field after this run.

216 - 2020-03-20, 12:51 PM

Again 196+ run, 29EE(so above avg), 0 leaks since 140, except 1 in 180.
But i looked at time after 197 : 22464 . it's even faster than WGW fastest 196 time currently on list(ground wawes now takes around 200s, so after 196 it was around 22,2k - still a bit slower than WGW fastest ever time).

i got perf. saphirre this run at wawe 10 , but created star at 18, and i didn't get BS before 30(as i remember), which speed up early wawes so much.
In addition got 1st BO after 100, and GO in 45 and i didn't have perf.o. before(untill 45 only normal opals was buffing my gems, strange that only 4 leaks in 44). Only Great gem that i have now is GO, i could take G. Diamond 3 times(no, thanks) and once G. saphirre(but outside of GO range).

So it's possible to do it faster.
i'm a bit suprised that emeralds runs can be faster than 196 runs with leaking all wawes after 160.

Anyone else is trying to do 196+ runs?

The only thing that is really needed to do this is GO and a bit of luck.(maybe is possible to beat 196 without GO, but it's extremelly unlikely)

I rarely leaking 3 with GO + BO, only if really bad EE gain, so chance for stack around 100 is very low.
The most dangerous are around wawes 120-150, where required distance between flyer to buy lives quickly increase. around 170 i rarely leaking 2(well, i died twice in 160, after 0 leaks in few previous wawes)

My previous 196+ run prove that very good EE gain is not requied.
chance for being able to take perf.e.(incuding combining gems) per pick is around 7.35%.
12 jades give gold for 190+ repicks, 196+190-19=367 assuming 370 picks with max chances
avg emerald gain is 27.2 (i succesed with 21).
chance for specific amount of emeralds (370 picks):

10 0.006%
11 0.016%
12 0.039%
13 0.084%
14 0.170%
15 0.321%
16 0.564%
17 0.932%
18 1.451%
19 2.132%
20 2.968%
21 3.924%
22 4.939%
23 5.928%
24 6.800%
25 7.466%
26 7.859%
27 7.943%
28 7.719%
29 7.222%
30 6.512%
31 5.666%
32 4.762%
33 3.869%
34 3.042%
35 2.317%
36 1.710%
37 1.224%
38 0.851%
39 0.575%

Note: if you get many emeralds you probably build less jades, so also lees repicks, so chances for 30+ EE are actually lower.

Rupert The Great - 2020-03-15, 01:34 AM

216 - 2020-03-14, 03:18 PM

216 - 2020-03-14, 02:58 PM

Rupert The Great - 2020-03-13, 03:03 AM

216 - 2020-03-12, 03:58 PM

216 - 2020-03-12, 11:45 AM

216 - 2020-03-11, 07:08 AM

I very like GTs, they are good. but GE sucks , i usually downgrade it to perf.e.(which are needed). Unless is in good position for slow. But Great gems(except 1st GO) almost doesn't matter at 196 - amount/positions if EE if more important, and most dmg is done by EGs/EEs anyway.
But dmg gems can help around 100 - less leaks is always good.
GO is best? - GO is needed for that overpowered aura(maybe perf.o. aura should be stronger?). but i prefer 2x GA + 1GO that 2xGO + 1 GA. 
So my favorite Great gems are GT and GA (GO is more wanted anyway).

Ofc, before 20 we can get perf. gems only by combinations. Yes it was from 2 flawless(4 normal are very rare). Actually i didn't uprade EG before 22 so i should say rather G than EG, but it was killing enemies quite quick in 17 and 18 (14 kills after that 2 wawes).
But not Smashing enemies as early BS. 
Once i got BS in 15, it was fun. Fun fact: i got it from 3 flawless rubies(20% flawless chance) - i placed fist 2 in bad place for BS. i had norm.a and flawless aq. eariel.

in one run (probably that failed in 39) i got perf.e in 12, buti could get shaphire for EE for long time, it wasnt't fun. p.e. got 70+ kills
i hate that when you combine special that kills are lost. Maybe specials should get sum of kills of their components.

Rupert The Great - 2020-03-11, 05:38 AM

216 - 2020-03-10, 10:40 PM

So how many leaks/EE? without GO game probably will fail anyway.
You probably have more emerald runs experience than WGW, because he is mostly doing 196 runs(i also think he should do past 196 run at leat once). 
I'm interested who will be 3rd person that submit legit 196+ lvl score(also want to talk about 196+ game balance, currently nobody else experienced problem that it's hard to lose after builing many EE?).
However beating 196 requies patience(and ability to not destroy your computer in anger after failing good run).
However,with some experience, GO emeralds runs have much bigger chance for succes than i exepted.

i'm testing new maze - even shorther and very optimal for space for gems.
i died in 38 once - no BS, no perf.o., gols used for repick ofc (but 3+1EE so shame),
also in 39 once - removed rock and forgot to replace.

i'm in 70 now - got EG in 17(eariest ever,164 kills now), leaked 1 in 39, but bad luck before 40(except that EG).
Also got 1-turn paraiba few lvls ago(about 1:12k chance, last time i got it 1-2 months ago).

i'm always thinking why your time scores are so bad (very long mazes?).

Rupert The Great - 2020-03-10, 09:45 PM

216 - 2020-03-10, 07:10 AM

*tab, not window, anyway you did similar mistake to me :(
Did you had a GO?
Ctrl+F3 is bad key combination for auto click.(even unlikely, but possible to accidentally press alt+F4).

I yesterday died in 120 with GO, due to unlucky stack :(
But EE and EG gain was much below avg. 9+1 EE only. 11 jades(planned 12th). leaking 2 in 104,112,116, 0 in 108.

So my defeating 196 rate on GO runs drop to 50% :(

But last(closed browser) run almost couldn't be failed before 200(i didn't count that as run).
Also a week ago i got GO in 104 and instatnly died(perfect stack)
But i got GO in bad place, i didn't know i still can do that mistake(i always stop placing gems after getting perf.o, and place on mid), maybe lack of sleep or i lost hope in that run(generally everything was bad), but i was hunting especially for perf.o. (planned do BO after that), got them on last 2 gems.

But i need to count that as wasted GO run, 
so 4/8 succesed(actually started counting when defeated 196 1st time, so maybe it should be even lower?)

Rupert The Great - 2020-03-10, 05:24 AM

Rupert The Great - 2020-03-10, 05:22 AM

216 - 2020-03-08, 11:53 AM

lement - 2020-03-08, 11:05 AM

216 - 2020-03-06, 06:10 PM

I know that i can upgrade gem before 16. i was doing it few years ago with silver(i was noob, and didn't know any strat, but i got to 100+).
But in emerald runs that 3 chances less is significant. maybe it's only 0.2 EE less and almost 0.5% chance for GO(i got GO once just after last chance upgrade, in my unlucky failied 160 run), but you want to do some special gems fast, like EG and BS.
The only real reason of doing it is letting malate gain more kills before 20.

How many kill can you get on initial malachite after 19 wawes? (if single malachite run)

I only about 120 ussually.

wawe 20 with our RC ofc can be done, no perf gems even needed, just flawless opal for haste,some slowers, ametyst gems.

i don't build 2hxN lane, i do 1hxN horizontal line just above c6. ofc it may be not completed/contain few other gems, but it works well.

ofc before 196 you can do many lines; to beat 196 quite safe 26 EE(good positioned, and GO ofc) are needed. it can be done with less(once i did with 24)but with more risk.
Getting 26 perf.e. before 196 don't requies unusual luck - it just average gain if about 175 repicks.
10-11 jades give enough gold for this.
Make sure you have enough space for EE.

EG armor reduction last around 5s.
But if you have EG near EEs and they will stun in EG range it kepp hitting that target(in that situation on screenshot).

Good luck with getting GO!

lement - 2020-03-06, 04:51 PM

Starting 2 games at same time is a nice idea; will replicate. Also means I don't need to choose between "stun flyers" & "don't stun flyers" strategies XD

For avoiding death in 16/20: You can max out your early malachite instead. This means you'll get less "perfect possible" turns but removes pressure from "red crystal or die" and also puts more kills on it - the initial malachite can hold without RC waves 20 and 24 easily, and 28 with some support.

(Though current game did manage to get past 20 without RC or maxing thanks to early perfect emerald.)

Topleft corner EG: I've noticed that EG armor reduction lasts for a bit; i.e. if they hit enemy on c2 they'll stay reduced past midpoint. If I have the fliers going past the two blood stones, I want all of them have minimum armor. Still, there are better places to accomplish this goal from.

Emeralds: Personally I have prioritized the upper left 2w×N lane over bottom half 2h×N lane.

Advantages:

More time to kill before fliers come for second pass, ground units in range sooner, ground units get hit by the gems between c1-c5 when going over.

Disadvantages:

An enemy near c1 let go there will arrive near the center at about same time an enemy from c5 will; doesn't help as much center killbox with stun.

My previous killboxes have placed malachite about 1 above the upper BS, but with starting with it near the center I guess the bottom lanes make more sense.

216 - 2020-03-06, 07:44 AM

holding off upgrades:
I also often do it in first 4 wawes for next(1-move) malachite/silver.
But they are not nessesary, and i recently get angry when needed constantly reset game due to getting useless 5 chipped gems(that's why i start 2 games as the same time,in 2 windows), so i upgrage at wawe 3.

I often die in 12,16,20. Main reasons: no opals, no RC. I hate when i get emerald in 20 and die(unupgraded DE isn't that strong, maybe sometimes worth wait with combine special).

Generaly having a bit longer maze early and shorthenig it later is good idea, i ofter do that(actually i remove only few gems, my mazes are generally short), especially aroung 150, when you (should)have many emeralds, so ground unit don't get far.

13+3 EGs in 116?
Actually there is no limit of EGs, but if getting that much, i rather use more repick and igrone some ametyst/downgrade to flawless. Buffed EG is best sinle-target dmg gem(ok, GAm is better, but you don't get them so often, also GAq has a bit more dps, but is IMO worse(range), and uber stone win with all ofc),
but they never will be as important as EE.

that 5 rocks with 2 EGs are on right-vertical line which is good for EE(ofc not all of this line should br EE, i don't say making EG there was mistake, myabe upper better was EE)

that EG in upper lefl corner was for reducing armor only, i guess?
I don't build EGs in that places, because EGs have good dmg and i want have it buffed by opals.

Full buffed malachite have a bit more DPS that U235, but their slow aura is nice, sadly range is bugged. 1 should be made i think.
I hope in new HTML5 version it will be fixed (as well as not affecting by opals).

the lower of your 2 BS is in better AoE position IMO, at least for flyers. It range corner, which is important- enemies are closer to each other.

EE between c1 and c5 maybe  wasn't bad idea but it is rather for late game. first positions should downleft(that range 2 fragments), down right(even 3 fragments), around that 2 EGs, around your BSs - that also one reson do build them 1/2 tile moved right, so you can build EE on left(or reverse).

lement - 2020-03-06, 02:45 AM

Oh wow. That bit about jade definitely changes my strategy. No more making mazes, holding off upgrades, avoiding getting higher tier gems with deliberate goals to maximize jade (and minimizing malachite) kills.

Gem analysis:

Yeah I was forced to build 2 malachites from chipped parts :/ I formerly used to go for star ruby in center instead, but didn't know about the lack of multipliers. I try to hold off upgrading a bit for lucky special chances - while I don't like the lag of Malachite, I've figured the chance of 1-turn malachite/silver, less wasted kills and less deaths in 12/16/20 is overall worth it. Breakeven point for first is 4,47 vs EG if armor is covered, 3,65 for second. Used to go for star ruby as well, but that's when I figured "of course multipliers work on it".

Uranium also loses to buffed Malachite in damage while taking up perfect gem turn, so unless I'm going to place that on exit (or 1-turn it) I may well downgrade flawed opal to chipped. Though I dislike the lag induced by malachite and BS.

Silver however just plain loses to SYS as a slower and damager both. On that note, good observation on its damage (also went and checked sandbox that alternating SYS-MBO-SYS-BO the opal always overrides the 5% damage boost and the 5% damage boost never overrides the 40%). I've previously avoided them, but if I read it right might be preferable to build some normally over EG in upper left quadrant once armor is taken care of and disregard bloodstones past the best AoE points. On that map I placed 1 BS in center and 1 ABS near center, but I'm now thinking whether SYS would have been better instead.

The perfect opals were quite well placed, but probably should have taken second perfect opal turn on lower and then made a BO.

Emerald locations:

Yeah the 2 highest can both shoot c1 and c5; in c1 the bats go to rightmost side and then do 180° to turn back. If the rightmost highest was 1 tile below it still could, but no more than that (I intended to put a lucky pick star sapphire on the rock closest to c1).

With the most downleft, I deliberately didn't put it too close to the path so that it wouldn't be able to cover same path for long. Intended to go to the 3w×2h space on the right with the 2 downmost, but perfect opal didn't cover that space. If I could have decided the placements, they'd have been left for later, as placed smack in the middle of road they'll cover same bat for long.

Same idea of avoiding covering too much of path being why there's 2 EGs on the right - at that five-rock line, EE would hit the c2-c3 line and c3-c4 line in addition to c5-c6 line (intended to place 3 above that eventually, but never got the chance). Same reason why the blood stones are not 1 to the right.  On that note, I think 1 more square too low is too low; the upper BS location seems best possible place for BS to me.

The midmost was indeed somewhat forced (aimed for great, got flawed topaz).

Maze: You're right about some shortening; at least the hook on the lower right - it served in early waves, but no longer necessary until filled in. Not sure about middle right but might be profitable to maximize the exposure % to multi-target gems in center. Dwarves reached 5th (6th before).

Later waves also made me think of in the future removing one of the center teeth in favour of more gems.

Though overall I probably will use a different maze next time; the few first 20 times I've done haven't been terribly good time-wise, and the first 40 times, while ok, haven't been terribly special either. The C1-3-5-E separation is natural, but the C3-C4-C5 route ends up being a large timewaster; I've also skipped on separating the exit from start before.

Poison: Yeah, I've noticed that order of operations matters here (has been more annoying when chipped emerald conflicts with jade I've been building kills on). Why my maze has sapphire and EG in upper left, but not jade itself.

Armor: I noticed full armor taking about half damage down, but don't think I've noted down that exact formula.

PS: And looks like my run took the EGs :p, I think when I died I had 13 with 3 more perfect amethysts waiting.

216 - 2020-03-05, 09:03 PM

I noticed that jade gold proc is connected with stun proc, i trought they have separare rolls. Also gold gain is rounded up, in wawe 37 i was getting 19 gold (from lvl 0 jades, but i'm 99,9% sure it doesn't matter).

In your run(at lesat until 99) you didn't do any yellow star - looks like it was needed between c1 and c5.
Also dmg from this isn't that bad, EG have almost 3x more single-taget dmg, but due to huge splash start usually hit 3 or more units.
You also ignored uranium - maybe it can be ignored, but i ussualy do 1 around mid, for slow.
Ofc in Emeralds run we build mainly EE/EGs and maybe few ABS later, but don't ingrore completly others. Except for Pink Diamond - why it even exist(IMO it should have bonus aura that do something, maybe bonus range, bonus crit or reduce minimum speed - game would be more interesting if it wasn't useless).

Maybe you should add description of the table on wiki, that says those numbers are dmg multipliers, and opal=Aq/O.

I hate that emerald poison can be overwritten by jade poison, i remember i did GE between c1 and c5 fro slow(sacrifice 1 EE), and the jade nearby was overwritting poison, so it last 4s, instead 30s.

Also i dislike that 2 flawless and 1 perfect in the same color cannot be combined into great. Even more frustrating is getting 3 flawlees and 1 perf, because if the perf. gem was worse(flawless),you could make great gem.

idk if armor function is known here or you believed me that 1 is x0.9434, but i'm 99% sure that is:

1-0.06A/(1+0.06*abs(A))
it extractly match all armor values which i tested(1,8,12,13,14,15,on negative armor it also works, once i did Great Diamond on sanbox easy wawe 1 for wathing crits for 23k in wawe 39).
Similar function i found on reddit(but it's dota2 focused, i play only flash version).

216 - 2020-03-05, 06:40 PM

The Jade "gain half the current lvl gold" means game lvl (wawe) in my observations, jade lvl don't affect it all(i think). So in wawe 80(example) all jades have 1% chance for 40 gold, and i observed that many times. 
If jades gives 1/2 of their lvls it would be hard to gain gold at all, but it would be more interesting if jade lvl increase gold gain.

Your maze is interestng, but way too long, i only separate c1,c3,c5,exit from start,c2,c4,c6 and build some maze around initial malachite.
Single ground units should reach c6 (before 1st BO), if not the maze is definitelly too long. many important positions for EE end other gems was taken by your maze.

About positions of that 7 EE, which i can see:
That most downfelf is ok(should be even more near).
That 2 lowest are ok, but they should be even 1/2 sqare lower(if you want follow my horizontal line, which is stongest point of my start, i think).
That 2 higest are too high,i think , and 2 nd is 1/2 sqare moved right than could shot c1-c2 line(it had c1 in range?)
That one below is ok,
the near mid is generally not the best, but 1 mid EE at early wawe works well, so is acceptable(i also sometimes build 1st in mid).
The total EE that range mid point should be 3-5(this also apply to some other points of path).

4 malachites?
I calculated all gems on your map and the number was too low!(97, i multipled paraiba x4 and specials x3). the explaination would be you did malachite(s) from 3 parts...(NEVER do it, its waste of turns, get it in 1-turn or not at all, if i'm forced to take chipped gem, i reset the game).

anyway wawe 116(i know that scores shows 1 lvl less, but i prefer to say "lose in 116" than 115 wawes completed) is good as no GO(and BO) and general bad luck. you should do BO, but i know problem - need perfect opal to replace old, before BO.

i got today to 152 with no GO(new no GO record, previous 136), and general bad luck - from wawe 50-60 to 150 i got only 3 perf.e!! with around 100 repicks! Assuming 180 wawes+r it's less than 0.05% chance for that. i got next in 152, just before die. at around 50-60 i had 7 EE(good), but overall it was bad luck(with around 1% chance). In avg i should have 18EE at this point(wawe 150+120) so almost twice as much. Also bad EG gain as start...(when i had 7EE, i had only 1 EG, EG are impotant at stat, they need to get many kills).
Leaking 3-4 ussually,once 2 around 130, 5 in 148. However flyers didn't get stack, maybe distribution of
was good(0 in upper part of map). Always when group of flyers on c5-c6 line get to around to mid, EE separated them.

your positions of BS are similar to mine(mine are moved by 1/2 sq left/right, because of my maze. also like position 1 sq moved down, so 0.5 sq above path). One of my BS in 152 run had 120+ kills.

lement - 2020-03-05, 04:24 PM

Here went another emerald attempt at 99: 
https://i.imgur.com/UGh2gZV.png

Could have done better on great topaz and aquamarine placement (and had a second red crystal) if I had considered the possibility that they'd happen. Still, no GO or BO & way too little EE and relatively lot of gold - though second most jade has only 33 kills, third and fourth 10 each and remaining two are level 0 - in my experience, third+ jades have serious trouble leveling, so I question how much gold most of your jades contribute with their 0/200 gold per hit.

Ultimately, died at 115 from a stack (wish one could see the layout from high score), with 40k gold in stock, still without any GO or BO; leaked 4 on previous wave, 6 in 100. I need to spend more heavily, my repick cost was only like 1200 :/

The damage conversion table is interesting (and now on wiki) - I would have expected a colour wheel. Instead, only Amethyst and Ruby have secondary weaknesses at all, and other strengths are not so crippling either.

Using that table, calced the total damage done for first 40 and 21-40 rounds:
		1-40	21-40
Green: 43,6	21
Violet: 43,4	20,75	(17,5|8,75 for air only)
Blue: 42,4	21,2
Red: 39,4	20,7
Yellow 38,3	18,3
Cyan: 37,5	18,2
White: 31,5 	17,1	(actually lower by 2/1, given no white gem can hit flyers)

216 - 2020-03-03, 06:14 AM

216 - 2020-03-02, 07:57 PM

i tried to confirm your theory that below 5 speed unit can take 2 ticks. but to confirm that need to notice over 5 speed unit that take 0 ticks. I noticed 6.92 unit that got 0, 3.14 that got 1 and 4.4 that got 1. i noticed(twice) around 6 speed unit got 2 tick - it got 3 tick from 2 flamestikes. but i might was 3 - ABS can regenate mana 2.5s after  flamestike. so i'm not 100% sure.

about closing browser - i probably did ctrl-shift-w(it close 1 window in chrome) twice.  
I wanted to write "EE" in comment,so shift+e,maybe it was horrible missclick.

i think i finished dmg conversion table:

dw	a	e	t	o	s	r	d
a	1.75	0.8*	0.8*	0.8*	0.8*	0.8*	0.8*
e	1.5	1.7	0.7	0.7	0.7	0.7	0.7
t	1.2	0.7	1.6	1	1	0.55	0.6
o	1	0.7	1	1.9	0.4	1	0.8
s	1	1	1	1	1.6	1	1
r	0.8	0.5	1	1	1	1.8	1
d	0.2	0.6	1	1	0.75	1.25	1.6

but i have problem with ametyst dmg.(almost) only gem that hit ground unit with A dmg is EG(upraded or not). they have significantly difference between min and max dmg.
some of those values can be 0.75 instead 0.8.
The solution would be geting Uber Stone, which have const dmg. but we know chance, i have never got it.
Anyone can confirm that values?

Jade text is scoling in sandox well for me.

I can see on wiki duration of frost nova, but it dosen't matter, enemy which enter in nova don't getting dmg(like in flamestrike). Nova can affect few targets, but only instantly.

lement - 2020-03-02, 06:51 PM

216 - 2020-03-02, 01:47 PM

lement - 2020-03-02, 01:39 PM

216 - 2020-03-02, 01:23 PM

lement - 2020-03-02, 12:48 PM

216 - 2020-03-02, 12:34 PM

lement - 2020-03-02, 11:47 AM

216 - 2020-03-02, 11:04 AM

lement - 2020-03-02, 09:43 AM

Well done 216, and that's a good point about star sapphires - I've used silvers before, with them covering c1 and c5, but at higer levels they'll fail to have enough hit rate/splash in their normal state to cover everything all the time.

Certainly motivating to do another emerald run, especially as I've failed with autoclicker on lower left corner run before despite bats being spread out perfectly. (That they're much more enjoyable in speed helps.)

The exact multipliers on damage values are also welcome; I also noticed green damage doing more than could be expected from just-amethyst weakness but didn't find exact values in sandbox.

Unupgraded Bloodstone splash range is still unknown though, if it hits adjacent even on high levels it'd do 135*(3*3+2*2)*0,9434*0,8=1324 with 5 enemies on 1 armor.

Flamestrikes ignoring armor and general under half multiplier makes sense of their weird behaviour vis-a-vis kills, good to know they ignore black opal too.

@Joaquin Phoenix: Remind me, how did you scroll Jade description? I've done it before, but have forgotten it.

216 - 2020-03-01, 05:31 PM

Leaking 3-4 with GO around 100 is much, i very rarely leak 3, ussually 1, almost no matter how many EE i have at this point.
Maybe enemies wasn't slowed property? it's big diffence between 6.92 and 4.91 or 5.53 speed, yellow shaphires help a lot with slowing many targets. You had 2 silvers which are similar.
Or you take non-great opal that slow your gems?
Or all EG's was hitting the same target(how many you got?)

Maybe you didn't have EE in important positions. Imagine that you place gem just above c6(touching c6). then one move it max 3,5 squre left. The next one move max 3,5 sq right. IMO those are the best positons, left one can hit fragment after c2 ,then swap to enemies before c6. the right one can hit even 3 fragments.

anyway bad luck - perfect stack aren't happens so often. At 104 only 1 click is needed between flyers(if 50 lives).
I noticed that spreding single EE increase chance to stack.
They can't stun ememies to long - usually 1 stun and enemy escapes.
So better is making groups of 2-3.

i also lost a run with GO in 104, few weeks ago, because of turned off auto clicked (i start clicking on mouse button 5, but before it i need to change layer - i want to use this button normally when not playing GTD on late wawes, i have almost no leaks before , so trought it's no needed auto clicker yet,"what bad can happen?". But i leaked 2 extremelly close. When i noticed this i tried to hit key combination that swith mose layer, but i missed, i had no time, so quickly stated clicking manually, but failed)

Joaquin Phoenix - 2020-03-01, 04:12 PM

216 - 2020-03-01, 03:07 PM

Joaquin Phoenix - 2020-03-01, 02:36 PM

216 - 2020-03-01, 05:33 AM

Joaquin Phoenix - 2020-02-29, 09:41 PM

216 - 2020-02-29, 07:24 PM

216 - 2020-02-29, 02:37 PM

Joaquin Phoenix - 2020-02-29, 02:21 PM

216 - 2020-02-29, 01:41 PM

Joaquin Phoenix - 2020-02-29, 11:59 AM

Joaquin Phoenix - 2020-02-29, 11:56 AM

216 - 2020-02-29, 11:50 AM

Joaquin Phoenix - 2020-02-29, 11:29 AM

216 - 2020-02-29, 08:20 AM

Definitely it is, text between "less" and "greater" was removed(i wrote " is less than 20%, not more than 20"). 
Also your highscore with "Got great opal on LAST TURN" still exist on list.
Times on (if doing property) emeralds run aren't so bad, because high dmg, and grouns units get stunned in good dps zones and short maze. You can see i had 17k s on 160. Around 120 it was 11k. if not fail it would be 23-24k at 196. Similar to my first 196+ run (24,5k).

I have never liked your strategy - i prefer killing units instead leaking them (which is objective of TD games :), well late ground wawes are boring - forcing afk for few minutes, so idea for removing maze is good for your strat. I don't think i'll get better time than you while passing 196(but i was closer than i excepted), unless i'll get an Uber Stone, but we know chance(actually it 's nedeed to get it quite early, because it need to get kills to be stong). But there should be penatly for leaking many flyers in 1 wawe.

If you do 40 repick at 196 you still have 196+40-19=217 chances, so around 28% GO chance. I don't think that experienced player as you can waste chance for/ don't notice GO. Maybe you just got unlucky recently.

Are you upgrading chances after wawe 19 or you use gold for malachite uprade first?
it only few more chances, but i my 160 i wouldn't get GO i don't upgrade chances first(no way i'll survive to 159, where i got next).

you'll try to do emerald runs?
you said you had run with 25 EE, GO and GA(is not nedeed, just nice to have), but fail on 184, because you concentrated EE too much.
25 EE and GO in mid is definitelly enoght to attempt wawe 196(and you could build even more in next 12 wawes).
ofc, emerald must be property located. I suggest test it in sanbox(start with BS or GT to destroy faster ground wawes, also having stong BS is fun)
How many jades did you usually build in your emerald runs?

btw , got GO while writing this comment(lol, so much luck recently), wawe 53 now leaks:
44 2
48 3(one flyer with 8 hp, yes literally 8...)
52 1

so another big chance to beat 196 (4 EE already).

WGW - 2020-02-29, 01:29 AM

WGW - 2020-02-29, 01:17 AM

WGW - 2020-02-29, 01:16 AM

@216:  First off, congrats on a truly amazing score-- this game has existed for well over a decade, and no one-- NO ONE-- else has even come CLOSE to your recent top score.  :)  I think I'm more or less resigned, though, to sticking to my usual strategy, just because it's reliable and I like posting new level 195 scores (though I don't think I've managed a new level 195 score in weeks, for one frustrating reason or another.  I just lost my last game after level 193 due to some sort of autoclick failure I can't explain. . . )  But I really appreciate your taking time out to explain how you managed this feat!

As for the odds on getting a great opal, I will need to take some time out to really dig into the math you provided for the probability calculation, but I think I understand at least on a gut level why my experience of the odds of getting a great opal feels so far off from whatever the actual probability is.  It boils down, I think, to the difference in strategy.  For all intents and purposes, any GTD run using my usual strategy is basically over by level 100 or so, at which point I'm pretty much guaranteed a level 195 run based on already placed gems, with additional gems (both type and placement) not really mattering very much, aside from avoiding any stun/slow gems within range of air troops.  Once I create the first Black Opal (again, around level 100 or so), the game is more or less over other than the tedium of playing it through to the inevitable loss at level 196.

Because my priority is to shorten the total time it takes to get to level 196 as much as I can, I'm very, very, very stingy with repicks, so that I can use the accumulated gold to "buy" my way past the final 40 or so levels (saves me around 8 to 10k seconds per level 195 game).  So I rarely use more than 40 repicks by level 100, and once I've built the first black opal around that point, I hardly ever repick again.  Around level 160, I've removed most of the maze and am letting all enemies escape, including ground troops, at which point the gems I place mean absolutely nothing.

So there have definitely been many games I remember where I've gotten a great opal after level 100, or even worse, after level 160, at which point they're so totally useless to me, I don't think of them as great opal runs (because getting the great opal has very little effect, if any, on my ultimate score if i get one that late).  I even posted a high score once (before the old scores were wiped) that said "WGW got great opal on LAST TURN", a game where I literlally picked a great opal on level 196.

My feeling is that it's the combination of (a) my limiting repicking so much, (b) the fact that I don't think of any game where I got a great opal after level 100 as a "real" great opal run, and (c) I don't pay any attention at all to what types of gems I place after I start leaking everything around level 160 or so.  This is what makes great opals "feel" a lot more rare to me than they probably are in reality.

Thanks again for taking the time to explain how you arrived at the odds of getting a great opal, and your general strategy-- and congrats again on what may well prove to be the highest score anyone will ever post for this particular version of GTD.  :)

216 - 2020-02-28, 11:05 AM

I had another interesting run, but fail at 160, leaking 2, one was slowed by shappire, the next one wasn't, so they got too close for 7 auto-cliks. Only 17 EE, but if didn't get that unlucky leak, and got few more EE, it would be easy 196+. Interesting was getting 1-turn jade bettwen wawes 6 and 7 (10% normal chance), and GO in wawe 20 (after 19), just after last chance upgrades - first 2 gems was perf.o., also one of remaining was perf.emerald, but in that rare situation i don't take emerald :). No early EG, first one around 35.
Until wawe 90(+40 repicks) i had only 3 EE, after that i got next 10 quite fast, then again almost no emeralds. I build 12 jades(max amount that i allow in my runs, 3 was 1-turn).
In 159 i got next GO(wasn't usefull). I had no one other Great gem.
Leaks:
44	2
48	3
52	4
56	2
60	2
64	2
68	4
72	3
76	4
80	4	
84	5	
88	5	
92	4 (3 EE)
96	2 (1st Black Opal)	
100	1
104	1
108	3
112	2
116	2
120	1
124	1
128	1
132	1
136	1
140	2
144	1 (2nd BO )
148	2
152	0 (16 EE only, good stun RNG)
156	0
160	2 and die

Math (216) - 2020-02-25, 09:30 PM

I decided to do some math:

explaination of 1:652 chance for Great Opal(in 1 turn):

you need 2+ perf.o. or 4+ flawless, there is 10% perfect chance and 8 types - each time you place gem there is 1:80 chance it will be perf.o.(everyone know that).
There is formula for getting (extractly) n succeses with A chance in k tries: 
A^n * (1-A)^(k-n) * combinations(k,n)
,so 
succes chance (A) = 1:80 = 0.0125
fail chance =              0.9875
combinations(5,2) = 10

0.0125^2 * 0.9875^3 * 10 = 0.0015046 (1:664),
but this is chance for extractly 2 perf.o., you need add chance for 3,4,5.
But there is faster way - calculate chance for 0 and 1.

0.9875^5 * 1=               0.9390431
0.9875^4 * 0.0125 * 5 =     0.0594331
1 - 0.9390431 - 0.0594331 = 0.0015238 (1:656)

chance for flawless opal = 3:80 = 0.0375

0.0375^4 * 0.9625 * 5 = 0.0000095 (not much, but i got that once, long time ago)
0.0375^5 * 1 =          0.0000001 (even a bit less, should be 074 instead 1, but let's keep the same rounding. You think it cannot happens? Well once i got 5 normal topazes... The same chance for that, such as for any of 24 flawed-flawless gems).

because you cannot get 4 flawless and 2 perf. in 1-tunr, you can just add those chances.
0.0015238 + 0.0000096 = 0.0015334 (1:652).

Intuitive way to estimate (dont use it on exams :)
Rupert said if you got perf.o in first gem (1:80 chance), you have 5%(4:80) for GO(i correct him for 4,9%, but 5 is rounded). if you got if as 2nd gem (it's(79/80):80 chance, but use 1:80 for easy calc.), you have 3:80, etc. So (4+3+2+1)/80/80 = 1:640, not bad estimation.

So each time you place 5 gems(with max upgraded chances) there is 1:652 for GO.

Chance for getting GO in multiple turn:
just caclulate  1- (chance for not getting GO)^ turns.
you got last chance upgrade after wawe 19 and ussually i save gold for this.
Examples:
until wawe 44, and 10 repicks used:  44-19+10 = 35 chances,
1-(651/652)^35 =  5.23%
wawe 70+ 20 r. = 10.33%
wawe 96+ 40 r. = 16.44%
wawe 110+ 50 r = 19.46% (that why i said around 20% chance before lose)
wawe 110+ 60 r = 20.69%
wawe 140+100 r = 28.77%
wawe 176+130 r = 35.63%
wawe 196+130 r = 37.58%
wawe 196+160 r = 40.38%
wawe 196+200 r = 43.94%

,so i was suprised that WGW says he gets GO so rarely in 196 runs, it should be statistically once per 3 games at least(depends of repicks, if less than 50, once per 4 games)

Uber Stone chance:
(0.0125^4 * 0.9875 * 5 + 0.0125^5) * 8(any color) = 1:1034343

Specyfic special gem 1-turn chance(with perfect gem, except paraiba):
1 perf.gem + 1 2nd part + 1 3rd part:
0.0125 * 0.0375^2 * 0.9125^2 (any other gem) * 10(combinations) * 6(permutations)=0.0008782
1+2+1:
0.0125 * 0.0375^3 * 0.9125* 5 *24/2 (you can't change order of the same gems)=0.0000361
1+1+2: the same, 
1+2+2:
0.0125 * 0.0375^4 *120/4 =0.0000007
1+3+1:
0.0125 * 0.0375^4 * 1 *120/6 = 0.0000005
1+1+3 same,
2+1+1:
0.0125^2 * 0.0375^2 * 0.9125 *5 *24/2=0.0000120
2+2+1:
0.0125^2 * 0.0375^3 * 120/4=0.0000002
2+1+2 same,
3+1+1:
0.0125^3 * 0.0375^2*120/6=0.0000001 (even less)

total: 0.000964 (1:1036)

Average DPS of some (lvl 0) gems:
EG:
180/0.75 * 1.3(avg crit) = 312
vs 1-armor ametyst weakness unit : 312 * 0.9434 *  1.75 = 515
buffed by great + black opal (x2.8)  : 1442

EE:
150/0.7*1.45 = 310.7
310*0.9434*1.5 = 439
*2.8 = 1231

ABS:
200/0.8*1.3=325
325 * 0.9434 * 0.8 = 245
*2.8 = 686 (+500*flamestrkes active, assuming 1.25 avg flamesrike active is 1311 DPS).

Star Yellow shapphire:
180/0.8=225
212
594 (but huge splash)

Paraiba:
225/0.75=300
283
792 (+frost nova, but mana allows for 1 nova/2.5s, so 100 dps only)

Uranium:
256 (not 260, its bugged. In adddition it dmg strange irregullary (in my observations))
256
256

Malachite:
81.8
115
324 (in all tagrets)

Silver:
150
141
396

Great Ametyst:
375
619
1733 ,the big problem of GA is fact this cannot get many kils.

Great Aqua:
560
528
1479 (too low for that range IMO, many Great gems should be stronger)

Great Topaz:
250
250*0.9434*1.2 = 283
792 (x7 targets)

Great Ruby:
280
211
591 (x3 targets)

Great Sapphire:
200
188
528 (this gem also sucks, but if i got it, i ussually take it anyway)

Great Opal:

180
169
475

Great Emerald:
250
353
990 (IMO poison should benefit from kills, but it doesn't)

(Lucky) jade:
180.7
170
477

Red Crystal:
140
232
650

216 - 2020-02-25, 07:10 AM

Cleetus - 2020-02-25, 01:20 AM

Rupert The Great - 2020-02-21, 09:55 PM

Rupert The Great - 2020-02-21, 09:31 PM

216 - 2020-02-21, 08:45 AM

216 - 2020-02-21, 08:42 AM

216 - 2020-02-21, 07:35 AM

I did it. removed few rocks in 405 and some 1,6M hp ground ememies finally leaks. Game needs balance after 196, as i said eariel. 
I got crazy luck in getting perf.e., but doing it is possible with average emerald gain. Just continue adding e.e after 196 and game can last forever. in 252 game i had only few more emeralds than statistically should be, and i could not lose, but i didn't want to play forever, so i stopped adding e.e.

@Rupert The Great: i think you can defeat 196, especially that you said you got to 196. Actually i never failed after 140, maybe i have better luck(leaking 2 sometimes). But make sure your emeralds can stun enemies property. The ememy which first appear on map have higher prioryty for towers, and you can't change it. So place emeralds rather in max distace from path, so they can range only short fragment of path, preventing too many emeralds to hit 1 target - let's 3-4 e.e. to focus on one ememy. There is some places which can be target by more emeralds(4-6 ,because those places are too good) - usually between ckeckpoint 2(c2) and mid, betwwen mid and c3, and in massive stun zones around the c6. i place a lot of e.e in spaces between c6 and c2, between c6 and c3 - they can range 2 different places. You know that e.e should be spreaded, so place some at upper part of map too(should be at least 1 stun zone between c3 and c4(max 3-4 e.e per 1 emeny here), between c4 and c5, and some in long line after c5), but remeber that speading single emeralds increase chance of stack, because they stun only for short time(ussually, stunning is actually very random). After my 140 fail, i don't place much of them in range 100 of mid, max 3-5.
Between mid and c6 many emeralds should range(6+), enemies shouldn't escaping from here, also after c6(most e.e here also range around c3, but you can add few "emergency" emeralds, which range only after c6, i ussually do this around/after 196, because not enough emeralds. i don't ussually have good stun zones before c2, i had some in this game(1-2 flyers getting killed here), but i didn't fell it help me so much(that don't means you can't build e.e here at all).

Remember about slowing enemies! jades, yellow star,uranium(bugged with slow range:(,as i said in older comment), perf.e.(temporary) are good. Best slower id Great Emerald, but ussualy i prefer take perf.e. anyway. Unless i got between c1 and c5 and don't have stunners around c1, so it can slow all 10 flyers at start. You want to kill enemies on c2-c3 line before other get close to mid again!

Good luck on next run with GO!

216 - 2020-02-21, 05:48 AM

20% was an eastimated chance for getting GO before level +-110 , when you ussually lose without GO due to stack. with GO(and black opal)i almost never leak 3 units (well, once i leaked 3 and die in 140 with GO and 17-18 e.e, and leaking 0 in 136, it's was my biggest fail before defeating 196 fist time - reason was probably too many emeralds had mid in range - 1 got stunned perfectly on mid and rest passes without stun, as i remember). with out GO i leaking 3-4, so chance to fail is much bigger, i have never past 136 without GO. Around 130 enemies don't need to be stacked in 1 plae to lose - if they are too close autocliked don't save you. SO in most cases you need GO before 110.

Chance for GO is 1:652 per placing 5 gems, so math is simple :  1- (651/652)^(lvl-19+repicks),
assuming you take last chance upgrade after lvl 19 (there is small chance for GO before 20, but only once in my life i gor GO from 4 flawless, let's ignore that, calculated 1:652 include that event).
until 110 i use usually +- 50 repicks,so 20% is good estimation.

I plan to lose intentionally after defeating wawe 404.

In 396, i got another dangerous situation(just while writting this comment), but remeber that i could build 20+ e.e at this point, so losing "normally" at this point is almost impossible.

in my 252 run i remember i had not much worse time than on your high scores(at wawe 196), 24k s probably.

I really encourage to do emeralds runs, instead 196 runs- maybe in 90+% cases you fail early, but it's better than have 100% chance to die at 196.

wawe 399, over 1 000 000 000 score already!

WGW - 2020-02-21, 04:18 AM

@216:  I'm with Rupert The Great in being nervous on your behalf about a power failure or browser crash before you get to post your current game's high score-- it would completely blow away the previous record, which was level 211, by a/aznist, who also got to level 203 twice.  But that's it.  No one else has ever legitimately beaten level 196 except a/aznist, and now you (his/her high scores were posted with the name "aznist", but there are many comments on this page from the same person under the name "a").

I saved a copy of the old high score list before the game's designer, Peter Holko, wiped it clean last year.  Here's a screenshot of a's legit scores past level 196, but note that ALL other high scores higher on the list than my best score are very obvious fakes-- it's apparently very easy to spam the high score list if someone really wants to, according to Peter, though I have no clue how people do it.  He blocked the IP address of the most frequent offender, though, so I've been very pleased to see no fake high scores are currently on the list.

https://www.dropbox.com/s/6klqxtpeetqkv0k/GTD%20top%20scores%2C%20pre-2019.jpg?dl=0

I'm genuinely impressed by your very smart approach to emerald runs, which is very clearly working for you better than anyone else has ever managed before!  The challenge I've set for myself to keep the game interesting, even though I know my strategy will ALWAYS lose at level 196, is to optimize runs to reduce the time it takes to get to level 196, which means a higher score.  The best score I posted on the now-wiped previous high score list is still my fastest level 195 run to date-- I came pretty close to beating it once on the current high score list, but without some exceptionally great luck, I'm not confident that I'll ever get to level 196 any faster than I did before the high scores were wiped!

One question for you, if you don't mind (actually I have tons of questions, but I'll stick to just one for now):  You mentioned a 20% chance of getting a great opal by level 196 and I'm curious how you calculated that (I've never been great with probability calculations).  Going purely by the empiricial evidence of my having played thousands and thousands of GTD runs, that number feels much higher than what I've experienced.  It seems to me that it's not possible to calculate the odds on getting a great opal without factoring in repicks, which are entirely dependent on how many jades you build.  So it feels to me like a fairly complex probability calculation to determine the odds of getting a great opal by level 195; but I would estimate the rate of getting a great opal by level 195, again based only on my actual experience playing the game rather than a more abstract, mathematical probability calculation, would put the odds of getting a great opal by level 195 is a lot closer to 1% (if not lower).  I've found picking a great opal to be fairly rare, even using a strategy where I build up enough gold from jades to generally be able to "skip" (allow all enemies to leak) all levels from 165 or so to level 195 (that's what gives my runs the biggest speed boost, by skipping as many late levels with gold as possible).  Part of what has put me off of emerald runs is specifically the need for a great opal, since the prevailing wisdom has been that it's probably not possible to beat level 196 without one (though this is not definitively proven by any means).

Anyhow, once again, I REALLY hope you don't lose your current game due to a crash or power failure-- I'm looking forward to see your score posted soon!

Rupert The Great - 2020-02-20, 08:17 PM

216 - 2020-02-20, 05:03 PM

216 - 2020-02-20, 08:17 AM

42 e.e. are stunning enemies good enough(ofc they can't keep 10 enemies stunned forever, but long enough to kill), even with small mistake in positioning(and i can make much more of them, but now i making only dmg gems now - ABSs and EGs), I said eariel i probably will have to lose intentionally, flying units have 400k+ hp only(grounds 1,2M), because linear growth(as i said eariel), my gems still are getting more kills(main EG have 441,many EG and e.e. have around 70, GO got 128, my fist jade 99...) and i still adding new ABSs(but i have problem with space for gems now), so i can kill fist few flyers, before anything reach mid after chepoint 5. Repicks costs almost 7k now(i have 200k+ gold lelt), it looks like a sanbox(but no 10k lives).
wawe 309 now.
I also trought my 252 run would be like that, but 31 e.e(better located a bit) wasn't enough, i stopped building them too fast... but i least i could submit score.
Generally defeating 196 is simillar to 200, 204, etc., so if build some more e.e after succesfully 196 attempt, and you won't die quick in few next flying wawes game will become endless.
i'm suprised that nobody did it in past 10+ years(maybe did, but cannot sumbit score :)

What is needed to beat 196 (excert some experience):
-not bad start(later it almost doesn't mater, but getting early EG helps around 100)

-Great Opal,before you die in 96-196 (about 20% chance, if you can survive longer is more). with that amount of emeralds which i got(it was below 1% chance for that amount!) i'm thinking about possibility of doing it without, but let say it's now theory. GO it's not only 1.3x more dps than Perf.O., it's also more efficienty for stunners.

- at least average emerald gain(more than 50% chance). Ofc more emeralds - more chance for succes. Lets say you can do 160 repick before 196. I estimated it's 7,4% chance for p.e in 5 gems(don't forget you can combine), so (196-19+160)*0.074= 24.9 , i managed to succes with 24 e.e, but i died in 204(with 26 e.e, as you can see in highscores).

- don't getting unlucky struck in 96+(sometimes even before, when 3 enemies are too close to each other). If you got GO and build black opal you shouldn't leak more than 2 around wawe 100, even with not much emeralds(but few are needed), because if you got less e.e, you probably have more other dmg gems, like EG. 
E.E. are more and more important later.

Anyway positioning of e.e is most important. I don't think i found optimal strat, i still trying to optimize(maybe writting "wawe 196 simulator" in C++ or other language would be a good idea).

@Rupert The Great: did you really think that RC reduce armor only to 1 unit, like EG?

Rupert The Great - 2020-02-19, 08:45 PM

216 - 2020-02-19, 07:05 PM

Rupert The Great - 2020-02-19, 06:53 PM

216 - 2020-02-19, 05:29 PM

Rupert The Great - 2020-02-19, 04:52 PM

216 - 2020-02-19, 01:21 PM

216 - 2020-02-19, 08:31 AM

After more than week of bad luck, i got GO again(in wawe 28, 80+ kills in 196), in addition to most lucky ever emerald gain(35 perf.e. before 196, much more than needed). So, i don't think i finish this run quickly. I leaked 1 in 168 with 27 e.e, so i definitely did mistake with positioning them (not enough stunners that range fragments of path after checkpoint 5, before mid and around checkpoint 4, so enemies was escaping too easy. Some important positions was taken by other gems, but amount of e.e fixed the problem anyway).
In this run i really had problem with too much emeralds :), because problem with getting other parts of e.e and i had too litle other gems, so my dmg was bad in general (also bad time, but beter than Rupert times, in 176 i had 25k seconds, are You doing too long maze?)

leaks:
44	0
48	3
52	3
56	3
60	3
64	4
68	3
72	3
76	4
80	3
84	3
88	4
92	1 ( 1st black opal created)
96	2
100	2
104	2
108	1
112	2
116	1
120	0
124	1
128	2
132	2* (maybe 1, i forgot to write)
136	1
140	0
144	0
148	0
152	1
156	2
160	0
164	0
168	1
172+	0

i'm currently in 211.
@Rupert The Great: you said once that you build 7 Red Crystals, what was a reason to do this? Six 1-turn RC?

Rupert The Great - 2020-02-13, 07:11 PM

eze - 2020-02-13, 06:14 PM

216 - 2020-02-12, 06:52 AM

@WGW thanks for explaination! I trought that order of placing gems matters in case of this bug(because bugs doesn't happens random). Another downside(IMO) of getting GO too early is that it will take too many kills(we want more kills on our egyptian golds(actually emeralds also have good dmg) or good aoe/multi target gems like Great Topaz(i didn't got them in my both 196+ runs, but i like them), maybe single-target dmg is more important in 196+ runs, because ussually you need kill first 2-3 enemies before other(s) gets close to checkpoint 6 to allow towers(emeralds) to switch target, keeping 10 ememies stunlocked for enough long is unlikely). 
About defeating 196 twice in short time: if you don't count resets/fails before wawe 40 it was literally next game after 252 run, i wasn't excepted anythig. Except getting e.gold in wawe 20, i had nothing good/unusual at first 100 wawes, i was thinking about ingoring it and starting new(only reason why i continue was good amount af kills on my egyptian gold), i was leaking a lot in 96-104(gj autocliker :), but in 106(?) i got GO, then i built black opal. I had only 8 emelads in 108, but i leaked only 1(maybe 2).
Also i got lucky of getting emeralds after 170-180(i had 18 around 180). i had 200k gold, so planned repick a lot, but even it wans't much needed(actually i think losing in 204 with 200k gold was a mistake). So afert few months of trying i successed twice within 5 days.
@Rupert The Great:
Math correction: if you got perf.o. in fist gem, you have 4 chances 1.25% ,so it's not 5% but 1-0.9875^4 = 4,9% . remeber you can get more than 2 perf.o.(i got 3 twice in my life i think, probalby on sanbox).

WGW - 2020-02-12, 12:04 AM

WGW - 2020-02-11, 11:57 PM

I should clarify what I meant about all opals being bugged, not just the black opals-- I didn't explain it properly at all.  What I meant, really, is that all the opals are bugged, from chipped to great opals, and Black Opals, though the bug exhibits itself somewhat differently with Black Opals.

With all opals other than the Black Opal (including Great Opals), if you place a gem within its range later, it will still get the speed buff from the existing opal.  And yes, if you place another inferior opal range within range of other gems ALREADY boosted by a higher opal, their buffs get bumped down to the lower opal you just placed-- this behavior is a bit hard to puzzle out, but it absolutely isn't random, inconsistent, or something that only happens "sometimes".  Here's how it works:

Let's say you have a jade built, and put a great opal within range of it, giving it a 50% speed buff.  Because the jade already existed on the board when you placed the great opal, it will never lose its full 50% spped boost, no matter how many inferior opals you place within range of the jade later.

But let's say you build a dark emerald a few turns AFTER placing the great opal.  The dark emerald, of course, gets the 50% speed boost from the already placed great opal.  BUT if you NOW put down, e.g., a normal opal within range of both the jade and the dark emerald, the jade's speed boost remains frozen at 50% (because it was already on the board when the great opal was plced), while only the dark emerald's speed boost will downgrade to the normal opal's boost of only 20%.  Basically, opals (other than the black opal) don't just boost the cooldown of gems, they also determine a sort of "floor" boost for already placed gems within range, which prevents later, inferior opals from affecting their existing speed boosts; On the other hand, any gem within range of an opal, but placed on the board AFTER that opal was placed, WILL have their speed boosts negatively affected by any inferior later opals placed within range of it.

This is just about the only downside to getting a great opal really early in the game-- you have to be very careful about where you put any opal from that point on, since most of your special gems will be built after the great opal, and therefore are vulnerable to having their 50% speed buff downgraded!  (Of course, once you get the inferior opal off the board by doing a "Combine Special" with it, any gems whose speed boosts were being negatively affected by the inferior opal will revert to the full 50% great opal boost.)

Somehow I doubt I made any of this any clearer.  :)

Rupert The Great - 2020-02-11, 08:34 PM

Rupert The Great - 2020-02-11, 07:50 PM

216 - 2020-02-11, 04:06 PM

a - 2020-02-11, 03:47 PM

216 - 2020-02-11, 02:35 PM

Joaquin Phoenix - 2020-02-11, 01:57 PM

216 - 2020-02-11, 01:02 PM

WGW - 2020-02-11, 12:46 PM

216 - 2020-02-11, 06:37 AM

Joaquin Phoenix - 2020-02-11, 12:45 AM

Joaquin Phoenix - 2020-02-10, 10:17 PM