My (very slow & steady) statistics from 2 July 2025 to 30 October 2025

[CameronHamill]

I thought I would post my Patchwork statistics in case anyone is interested. Yes they are very slow and steady but overall I'm fairly happy with them.

I aim to eventually reach a 50% Win Rate against Medium AI on the Patchwork App.

On BGA I hope to be able to reach and maintain perhaps a 300 ELO and a 50% Win Rate.

They are very modest goals I'm sure to many top players but I'll be personally happy to be a solid intermediate player. I'd like to maintain that. Patchwork is the first digitised board game I've played (and the first board game I've played in decades).

As a wild guess perhaps Hard AI on the App is a player of around 275 ELO on BGA? And Ultra Hard is maybe 350 ELO? I'm really not sure how high ELO players would rate the abilities of Hard and Ultrahard. A BGA member has estimated that Medium AI is possibly around a 100 ELO to 150 ELO.

Edit:

The Ultra Hard ELO level on the app is possibly much higher than 350. Unknown as to it's speculated ELO. Thank you to FrankJones for his comment.

My Patchwork Android App Statistics as at 30 October, 2025

I've been playing Patchwork on the App since 2 July, 2025

My app statistics:

Ranked games (against human opponents)

Played 20, Won 2, Lost 18

Current RP: 525
Highest RP: 548

Versus AI

Only easy-peasy (Lula) 191 points. Won 12, played 15 (80%)

Only easy (Hoo) 198 points. Won 17, played 29 (58%)

Only medium (Ute) 206 points. Won 180, played 500 (36%)

Only hard (Frankenknopf) 184 points. Won 1, played 52 (1%)

Only ultrahard (Uwe) 177 points Won 0, played 38 (0%)

My BGA statistics as at 30 October 2025

Playing on BGA since September 2, 2025

No Arena games played

82 regular games played, the vast majority Original and real time with fast game speed.

45 wins (55%) BGA ELO 228

[FrankJones]

Check my patchwork stats.

And, I get completely demolished by the Ultra-hard opponent on the app. So, I would say the Ultra-hard on the app is way higher than your estimate.

[CameronHamill]

Thanks very much for replying Frank.

I had a look at your stats and they are very solid. I wonder if Uwe Ultra Hard is maybe on par with the best Patchwork player on BGA? I will probably end up eating my words though. Lol.

It seems pretty hard to gauge the skill level between human and AI opponents. On an Abalone thread a member suggested it can't be done. I suppose there's no way of knowing really.

Your statistics are excellent!

[GiannisIsTheBeast]

Kind of has me curious… if I wasn’t so cheap I’d spend the $3.99. Looks like it gets under $2 on steam during sales so maybe I’ll grab it then lol…

Even 500 level ELO players on BGA make a lot of mistakes so if the AI is programmed to play optimally it could maybe get to 600 or 700 ELO. There are quite a few things to keep in mind to be very good… they could all most likely be programmed though to be honest.

[iLikeTrainz]

I'm not the best Patchwork player on BGA (Perhaps one of the best real time, maybe) but I'm a solid 800 rated player.

For science I will invest the money for the app and try on the hardest bot and will keep you updated here

[GiannisIsTheBeast]

Hmmmmm…. Well I decided to buy it and not sure I should have lol I have not won a game yet vs it and I have 810 ELO. I typically don’t play many too many games vs top players so generally my wins are the result of the other person messing up at some point. The AI doesn’t really mess up as far as I’ve seen. The only slight mistake was placing a 1 x 1 in their main area when they could have put it in a 2 x 1 spot since the 2 x 1 was gone. Perhaps it calculated all outcomes and decided it didn’t matter… hard to say.

Kind of frustrating as I feel even when I’m ahead I will get behind at some point. It seems to get the 7x7 most times as well as the 1x1s. I suppose if nothing else it is good training to face good players… arg…. Sigh…. Blah!!!

It seems to just almost be perfect at placing pieces. Many times people could win games but don’t leave space for a large valuable piece mostly because they tried for the 7x7 but didn’t get there. This basically never seems to be the case with the top AI as it always has room to place the good pieces. I think you essentially have to play perfect to beat it.

Edit: Finally beat it after 10 games. I think I should have won like 2 more but not used to having 1 bad move = death. I think the interface messes me up at times too as I’m not looking far enough ahead sometimes.

[GiannisIsTheBeast]

Beat it a few more times. The last game it did have a somewhat inefficient start. I slightly messed up at the end but still won by 2.

The other games I won were situations I should win… first patch cross and a 2 button first patch for me and it had 3 non-button patches to choose from.

[iLikeTrainz]

I played a few games - all ended in narrow losses.

Safe to say the engine overall probably beats the best players on BGA, especially when any time is applied.

I also have to say I dont like the app at all, as the view makes it a lot harder to plan things compared to the circular view of pieces here. Definitely impacts my playing strength

[FrankJones]

I might try it a few more times by getting out my actual physical game and setting up the pieces in a circle and setting up my board and the opponent board so I can see everything in one glance, which is not possible on the app, as noted in this thread. I'm not on the level of the two of you (yet ), but I know I can do better against uwe than I have thus far.

[GiannisIsTheBeast]

So far I think I’m something like 4 for 16. It’s hard to say exactly because there seems to be a glitch in the stats that if you concede a game to the AI, it somehow gives you one or two wins so it shows many more wins because I conceded several times. Would be sweet if BGA had an infinite win glitch like that. Also the AI pictures for Uwe and the 2nd hardest AI seem to be switched on the player select board. Maybe it’s just the Apple version but I’ve found these somewhat amusing. At least no glitches during actual gameplay.