Page 2 of 2

Re: The ELO system in TTR – Magic barrier or frustration factor?

Posted: 09 November 2026, 23:04
by Cournot
Vader300 wrote: ↑09 November 2026, 08:25 The ELO system works as intended as others have suggested. Luck is indeed a factor in TTR, and that's why the top ELO is capped in the 800s currently. Games where luck isn't a factor will naturally have a higher ELO cap (just look at games such as checkers and connect 4), and games that are highly dependent on luck will have a lower ELO cap. For example, the highest ELO in "can't stop" is in the 400s.

I think your post is particularly relevant to me because I recently did an experiment playing only players rated 100-300 to see how it will go (I usually only play 300+). I did not win every game, but I won enough games to which I didn't experience a net ELO loss compared to playing higher ranked players. At one point, I won 75 games in a row. Was luck on my side for all of those games? Absolutely not. But I was able to play my cards effectively in a lot of those games to overcome the difference in luck. This is not meant to belittle lower ranked players or sound arrogant. Just providing a supporting example. Of course, there are games where the luck difference is too great to overcome, but if you play your cards right, you can still win a lot of games where luck is against you.

Overtime, luck should favor both players evenly. So if you continuously feel like luck is against you, then I suggest reflecting at your own play. Often times what seems like luck is just great execution by your opponent.
Playing below 400 ELO of your own ELO is -EV btw.
The best is playing 500 to 800+ if you're skilled enough (which should ideally be the case if you want a ''max ELO'').
You will also earn more respect than farming sub300. My 2 cents.

I can't quote all relevant messages but I will add my points here to avoid replying too many times.

1) @Vader300 The top ranking is capped in the 800s for two reasons: Lack of skills (yes, you read me correctly) and the pool of 700+ players is not sufficient to ''support'' 850-925 (yet?). 8 players achieved 850+ so far [Flop Kasi 888 superblackbird 886 Knockando 885 admitted 880 scorps 871 Cournot 862 Sysyphus 857 Fleurdelysé 852]. Also, ELO-system is built to limit the difference between players to 400 pts ELO, which is not the case on BGA. Mathematically, this means that one can not have more than +/-92% as win probability (you are garanteed to win at least (1-0.92)*20=1.6 points per game won).

2) @MalteseFalcon62 The real problem here, when you have the perception of luck in your games against lower ranked opp, is that you lack sufficient game experience; i.e. that you cannot play not so obvious moves that you cannot find in ''Real Time'', only on ''Turn Based'' when you can think 5+ minutes per move. In turn based, it's almost impossible for a 700+ to lose against a sub300, regardless of colours or tix.

3) @Jellby Your ELO on TTR can have some wild swings. Most players don't improve a lot past their learning curve (truism). But their ELO will move because their level fluctuates, not because of natural oscillations. You should see TTR ELO as performance based.

4) @MalteseFalcon62 To complete my previous answer, the asymmetric risk push lower ranked players to take more risks, irrational ones, in the objective of outcome maximization. I don't like to discuss playstyles publicly, but let's say that I will more likely play GTO against a 700+ than a random on Arena (which is why I don't pursue Arena on USA). The pairing that you are describing is one of the reasons why Arena should not be taken seriously (sorry). It's true that skill can mitigate that, as others rightfully pointed out, but that is beside the point that ELO is made to compare players of ''about the same level''.

5) @Synthetase That is simply not true, you are always blessed by better tickets than I do.

6) @ExaltedAngel Excellent point, there should not be a 100 floor, but playing against sub300 should be forbidden to all 700+.

7) @Normal Distribution To go further, one who seek mastery should watch games at least as often as one plays. Better if it's live.

Re: The ELO system in TTR – Magic barrier or frustration factor?

Posted: 10 November 2026, 02:33
by Vader300
Cournot wrote: ↑09 November 2026, 23:04
Vader300 wrote: ↑09 November 2026, 08:25 The ELO system works as intended as others have suggested. Luck is indeed a factor in TTR, and that's why the top ELO is capped in the 800s currently. Games where luck isn't a factor will naturally have a higher ELO cap (just look at games such as checkers and connect 4), and games that are highly dependent on luck will have a lower ELO cap. For example, the highest ELO in "can't stop" is in the 400s.

I think your post is particularly relevant to me because I recently did an experiment playing only players rated 100-300 to see how it will go (I usually only play 300+). I did not win every game, but I won enough games to which I didn't experience a net ELO loss compared to playing higher ranked players. At one point, I won 75 games in a row. Was luck on my side for all of those games? Absolutely not. But I was able to play my cards effectively in a lot of those games to overcome the difference in luck. This is not meant to belittle lower ranked players or sound arrogant. Just providing a supporting example. Of course, there are games where the luck difference is too great to overcome, but if you play your cards right, you can still win a lot of games where luck is against you.

Overtime, luck should favor both players evenly. So if you continuously feel like luck is against you, then I suggest reflecting at your own play. Often times what seems like luck is just great execution by your opponent.
Playing below 400 ELO of your own ELO is -EV btw.
The best is playing 500 to 800+ if you're skilled enough (which should ideally be the case if you want a ''max ELO'').
You will also earn more respect than farming sub300. My 2 cents.

I can't quote all relevant messages but I will add my points here to avoid replying too many times.

1) @Vader300 The top ranking is capped in the 800s for two reasons: Lack of skills (yes, you read me correctly) and the pool of 700+ players is not sufficient to ''support'' 850-925 (yet?). 8 players achieved 850+ so far [Flop Kasi 888 superblackbird 886 Knockando 885 admitted 880 scorps 871 Cournot 862 Sysyphus 857 Fleurdelysé 852]. Also, ELO-system is built to limit the difference between players to 400 pts ELO, which is not the case on BGA. Mathematically, this means that one can not have more than +/-92% as win probability (you are garanteed to win at least (1-0.92)*20=1.6 points per game won).

2) @MalteseFalcon62 The real problem here, when you have the perception of luck in your games against lower ranked opp, is that you lack sufficient game experience; i.e. that you cannot play not so obvious moves that you cannot find in ''Real Time'', only on ''Turn Based'' when you can think 5+ minutes per move. In turn based, it's almost impossible for a 700+ to lose against a sub300, regardless of colours or tix.

3) @Jellby Your ELO on TTR can have some wild swings. Most players don't improve a lot past their learning curve (truism). But their ELO will move because their level fluctuates, not because of natural oscillations. You should see TTR ELO as performance based.

4) @MalteseFalcon62 To complete my previous answer, the asymmetric risk push lower ranked players to take more risks, irrational ones, in the objective of outcome maximization. I don't like to discuss playstyles publicly, but let's say that I will more likely play GTO against a 700+ than a random on Arena (which is why I don't pursue Arena on USA). The pairing that you are describing is one of the reasons why Arena should not be taken seriously (sorry). It's true that skill can mitigate that, as others rightfully pointed out, but that is beside the point that ELO is made to compare players of ''about the same level''.

5) @Synthetase That is simply not true, you are always blessed by better tickets than I do.

6) @ExaltedAngel Excellent point, there should not be a 100 floor, but playing against sub300 should be forbidden to all 700+.

7) @Normal Distribution To go further, one who seek mastery should watch games at least as often as one plays. Better if it's live.
I agree with you that having more skilled players will lead to a higher ELO cap, however TTR already has a very big pool of players compared to other games on BGA and luck still plays a major factor in determining ELO cap. Connect 4, for example, has a much smaller player base compared to TTR yet has a ELO cap well over 1000. And you'll notice this across games on BGA. The more the game is dependent on luck, the lower the ELO cap will be provided there is a sufficient player base to provide enough sample size. The ELO cap for TTR is in the 800s, as luck plays a role in the game but not as big of a role as a game like "can't stop".

And I played 100-300 precisely as an experiment, having always played 300+ before, and there is enough of a sample size to conclude that playing lower ranked players is not -EV for me compared to playing higher rank. I don't feel like playing higher or lower ranked players makes a significant difference on my net expected ELO gain over many games. I thought this was relevant as OP discussed playing lower ranked players who get lucky, because luck can be countered by strategy.

Re: The ELO system in TTR – Magic barrier or frustration factor?

Posted: 14 November 2026, 13:16
by MalteseFalcon62
Cournot wrote: ↑09 November 2026, 23:04 2) @MalteseFalcon62 The real problem here, when you have the perception of luck in your games against lower ranked opp, is that you lack sufficient game experience; i.e. that you cannot play not so obvious moves that you cannot find in ''Real Time'', only on ''Turn Based'' when you can think 5+ minutes per move. In turn based, it's almost impossible for a 700+ to lose against a sub300, regardless of colours or tix.

I respectfully disagree with the idea that perceiving luck as a significant factor is mainly a sign of insufficient experience.

A strong player can fully recognize the correct strategic choices in a game while also acknowledging that ticket draws, card distributions, or blocking opportunities had a major impact on the outcome. Skill and variance are separate concepts.

I agree that stronger players generally mitigate bad luck better than weaker players.
They find more flexible lines, adapt more efficiently, and convert advantageous situations more consistently. However, that does not imply that luck becomes irrelevant, nor that noticing its impact indicates a lack of experience.

I am also skeptical of the statement that a 700+ player will "almost never" lose to a sub-300 player in turn-based play, regardless of colors or tickets. That seems like a very strong claim. If colors and tickets truly had negligible influence, the stronger player’s win rate should approach 100%, which is not what Elo systems are designed to model.
A more balanced conclusion, in my view, is that increasing skill reduces the relative impact of variance, but it does not eliminate it. Recognizing variance and recognizing skill are not mutually exclusive.

Re: The ELO system in TTR – Magic barrier or frustration factor?

Posted: 14 November 2026, 14:21
by Cournot
MalteseFalcon62 wrote: ↑14 November 2026, 13:16
Cournot wrote: ↑09 November 2026, 23:04 2) @MalteseFalcon62 The real problem here, when you have the perception of luck in your games against lower ranked opp, is that you lack sufficient game experience; i.e. that you cannot play not so obvious moves that you cannot find in ''Real Time'', only on ''Turn Based'' when you can think 5+ minutes per move. In turn based, it's almost impossible for a 700+ to lose against a sub300, regardless of colours or tix.

I respectfully disagree with the idea that perceiving luck as a significant factor is mainly a sign of insufficient experience.

A strong player can fully recognize the correct strategic choices in a game while also acknowledging that ticket draws, card distributions, or blocking opportunities had a major impact on the outcome. Skill and variance are separate concepts.

I agree that stronger players generally mitigate bad luck better than weaker players.
They find more flexible lines, adapt more efficiently, and convert advantageous situations more consistently. However, that does not imply that luck becomes irrelevant, nor that noticing its impact indicates a lack of experience.

I am also skeptical of the statement that a 700+ player will "almost never" lose to a sub-300 player in turn-based play, regardless of colors or tickets. That seems like a very strong claim. If colors and tickets truly had negligible influence, the stronger player’s win rate should approach 100%, which is not what Elo systems are designed to model.
A more balanced conclusion, in my view, is that increasing skill reduces the relative impact of variance, but it does not eliminate it. Recognizing variance and recognizing skill are not mutually exclusive.
I will not debate against an AI.