Fix your ELO

[HRG_Surplus]

There is no point on having an ELO grading system. If your system keeps pairing players with high ELO like 300 to players round 75 to 120. There is little to no point on having an ELO system if this is a constant occurrence. If it's only there to create some value for purchasing the subscription plan on this platform. That is literally the wrong way of doing it. The point of ELO is to establish a ranking so the system can pair you up with other like ELO rating players.

It's a total waste in my opinion unless it's just a simple guideline to show if a person just wins a lot. In any other games that has a rating system of some sort. They actually pair players with similar ratings. Not a rating from low to extremely high in triple digits. That is insane to me. Or is this ELO rating just fluff? Thank you

[Phoxtrot]

Actually, that's not the issue.
ELO when matching people of different skills works fine for games of skills.
The problem is that the ELO system as is, doesn't work properly for games with a high luck component.
ELO system is meant to stabilize after a while so that differences in ELO reflects a difference in terms of chances to win.
And If I understand it correctly (my understanding of it is limited), it makes a bell curve where it is assumed that the best players in the world are pretty much certain to win against really poor players.

But take a game with a good amount of luck like Jump Drive and ELO never really stabilize or really reflects the chances to win, this is because once you have a little bit of experience, you get a reasonable chance to win against anyone. The best players will still win more often than lose because it's not a 100% luck game but put the best player in the world against a 150 ELO and he won't have a 90+% chance to win.
With the ELO rating as it is, you need more luck than skill to reach #1. I don't say that to complain about my situation: I had that luck and I did reach #1.
Also, nobody has expert status in that game despite the high number of players. Sure, we could cheese the system by inviting people to only play against players with the same ELO and at some point, someone would reach 500 but that is hardly fun nor a fix to the system.

To actually fix this, it requires non-trivial mathematics.
You need a ELOish system that is meant to represent a flatter curve of chance to win versus skill difference.

Suggestions to upvote (or not, it's your choice):
- https://boardgamearena.com/bug?id=14455

[Phoxtrot]

For some maths about it,
See for instance:

https://arxiv.org/pdf/2104.05422.pdf
https://gki.informatik.uni-freiburg.de/ ... Chance.odp

And some more accessible stuff:
https://medium.com/swlh/measuring-the-e ... 9a2612bd11

[Den_Man]

ELO is especially meaningless, because it is very common for someone to be a highly experienced player _before_ they start playing on BGA.
It's not fun for the person who takes the -30 ELO hit on that one. (Or even worse, that person really is a total beginner who has no idea how to play, and they get matched up against a 300+. It's certainly not fun for the new player.)

If it wasn't for ELO, I would happily play against new players. I'd help them out, give them tips, chat a little, and try dumb strategies on purpose. But if I do that now, I get punished for it.

[Phoxtrot]

ELO is especially meaningless, because it is very common for someone to be a highly experienced player _before_ they start playing on BGA.
It's not fun for the person who takes the -30 ELO hit on that one. (Or even worse, that person really is a total beginner who has no idea how to play, and they get matched up against a 300+. It's certainly not fun for the new player.)

If it wasn't for ELO, I would happily play against new players. I'd help them out, give them tips, chat a little, and try dumb strategies on purpose. But if I do that now, I get punished for it.

That's a problem a too, depending on how you select your opponents.
I don't really find that to be much of a problem for me once the game has been out for a while, especially as I mostly play arena mode.
But it is a shame in principle to have a system that pushes some of us into not playing with players who are new to the game.

I would be in favor of everyone starting with a more average ELO.
Another more complex idea would be that when you play against someone with less than 15 plays, your ELO score gets re-adjusted when that person reaches 15 plays. It wouldn't need to make perfect mathematical sense to help.

[Meeplelowda]

ELO is especially meaningless

If you truly believe ELO is meaningless, then why do you care about it enough to avoid new players. If their ELO is meaningless because they may be very experienced, yours is meaningless because your ELO loss against that player is unduly large because they were underrated.

I guess I find it a hard concept to understand why someone cares about something that in their own assessment is meaningless.

[HRG_Surplus]

It's meaningless for the ELO, since the main attraction for getting Premium is the rating portion of the service perk. But when ELO isn't effective then what is so special for getting the subscription and entering in tournaments. To me the company is looking for funding to provide a service that isn't technically working properly due to sheer luck. All board games is basically luck. Either getting the right card or rolling of dice. There is little to no skill in board gaming. The only part is understanding the game and different routes to winning. The ELO isn't the same like * since * is based on skill most entirely with some luck. But board games are pure luck. So if someone isn't lucky then there is the issue. The ELO should try and minimize that factor. But it doesn't, like when I am playing against a player that has a 200 point ELO count over me. The much higher ELO 90% wins like it's a guarantee. And that higher player than gets higher rankings through Arena and tournaments. It leaves the lower ELO players and makes it literally impossible to level up higher. It's almost a ripoff. If I didn't purchase the Premium service, I wouldn't have an issue with it. But I did figuring that this is legitimate practice. But now I am second guessing as it is fluff to help with funding for their platform. If I could get a refund I would. It's a worthless service in my opinion. Either fix it or not. I like the idea you start out higher and then it balances out.

What also irritates me is when I beat a higher ELO in a match with multiple players I get single digit gains. But if a higher ELO over 200 beats a lower level ELO they get a raise of double digits in the same position if I was in that position before. It's absurd that this is completely balance it's the polar opposite of being completely broken system just to get funding. It makes me not want to play. Because going into it I know I have no chance why even play a game. It's like playing 1 on 1 basketball with a Varsity basketball player and betting $20.00 on the game. I am for sure losing my $20.00 bet. This whole ELO makes no logical sense. And anything illogical is asinine to think it is worth to pay the fee for the ranking purposes and tracking. They need to get on this ASAP.

*Moderator Edit: advertising for other gaming platforms is not permitted on BGA.

[Tisaac]

It's meaningless for the ELO, since the main attraction for getting Premium is the rating portion of the service perk. But when ELO isn't effective then what is so special for getting the subscription and entering in tournaments. To me the company is looking for funding to provide a service that isn't technically working properly due to sheer luck. All board games is basically luck. Either getting the right card or rolling of dice. There is little to no skill in board gaming. The only part is understanding the game and different routes to winning. The ELO isn't the same like * since * is based on skill most entirely with some luck. But board games are pure luck. So if someone isn't lucky then there is the issue. The ELO should try and minimize that factor. But it doesn't, like when I am playing against a player that has a 200 point ELO count over me. The much higher ELO 90% wins like it's a guarantee. And that higher player than gets higher rankings through Arena and tournaments. It leaves the lower ELO players and makes it literally impossible to level up higher. It's almost a ripoff. If I didn't purchase the Premium service, I wouldn't have an issue with it. But I did figuring that this is legitimate practice. But now I am second guessing as it is fluff to help with funding for their platform. If I could get a refund I would. It's a worthless service in my opinion. Either fix it or not. I like the idea you start out higher and then it balances out.

What also irritates me is when I beat a higher ELO in a match with multiple players I get single digit gains. But if a higher ELO over 200 beats a lower level ELO they get a raise of double digits in the same position if I was in that position before. It's absurd that this is completely balance it's the polar opposite of being completely broken system just to get funding. It makes me not want to play. Because going into it I know I have no chance why even play a game. It's like playing 1 on 1 basketball with a Varsity basketball player and betting $20.00 on the game. I am for sure losing my $20.00 bet. This whole ELO makes no logical sense. And anything illogical is asinine to think it is worth to pay the fee for the ranking purposes and tracking. They need to get on this ASAP.

*Moderator Edit: advertising for other gaming platforms is not permitted on BGA.

And since when paying premium is only about that ??

I strongly suggest you to try games with less luck if you dont like that. There are many good games with very few luck involved and that's reflected in the fact that top players have very big ELO since they are able to consistently win (compare top ELO of Agricola with top ELO of Bang for instance)

[Blacktango]

Personally, I think the main advantage of being premium is that you can create tables for any game you want, and not have to wait for someone premium to create them.

Calculating ELO is something really hard, especially with so many different games on the platform.

I totally agree with what was said about the luck part, but we can see that for pure luck games, the higher ELO never reach high points.
If you take a game like the Yahtzee (pure dice game), the highest player has 4XX ELO. For a more strategic game like Race for The Galaxy, the top is close to 700 ELO.
We can see that the luck part is taken into account by the ELO calculation system.

IMO, ELO is just an indicator to help filter the players you want to play against/with, not a precise ladder to look at too closely, or a goal to reach.
Btw, another good indicator, to combine with the ELO, is the number of games the player has played.
If a player has 50 ELO, but has only played 1 game (because they won their first game against a 400 ELO player, for example), you can't assume that they are very experienced (maybe they played the game a lot out of BGA before, maybe it was just luck).

I think the Arena mode is meant to fix this. It defines a more specific version of the game (with a specific number of players, enabled extensions, …) which makes it possible to calculate the ladder more accurately. Here you can better compare yourself with other players, and try to be "the best".

The downside to this is that there will be less interaction between players, as some will play exclusively in arena mode and others in normal mode.

[Meeplelowda]

It's meaningless for the ELO, since the main attraction for getting Premium is the rating portion of the service perk. But when ELO isn't effective then what is so special for getting the subscription and entering in tournaments. To me the company is looking for funding to provide a service that isn't technically working properly due to sheer luck. All board games is basically luck. Either getting the right card or rolling of dice. There is little to no skill in board gaming. The only part is understanding the game and different routes to winning. The ELO isn't the same like * since * is based on skill most entirely with some luck. But board games are pure luck. So if someone isn't lucky then there is the issue. The ELO should try and minimize that factor. But it doesn't, like when I am playing against a player that has a 200 point ELO count over me. The much higher ELO 90% wins like it's a guarantee. And that higher player than gets higher rankings through Arena and tournaments. It leaves the lower ELO players and makes it literally impossible to level up higher. It's almost a ripoff. If I didn't purchase the Premium service, I wouldn't have an issue with it. But I did figuring that this is legitimate practice. But now I am second guessing as it is fluff to help with funding for their platform. If I could get a refund I would. It's a worthless service in my opinion. Either fix it or not. I like the idea you start out higher and then it balances out.

What also irritates me is when I beat a higher ELO in a match with multiple players I get single digit gains. But if a higher ELO over 200 beats a lower level ELO they get a raise of double digits in the same position if I was in that position before. It's absurd that this is completely balance it's the polar opposite of being completely broken system just to get funding. It makes me not want to play. Because going into it I know I have no chance why even play a game. It's like playing 1 on 1 basketball with a Varsity basketball player and betting $20.00 on the game. I am for sure losing my $20.00 bet. This whole ELO makes no logical sense. And anything illogical is asinine to think it is worth to pay the fee for the ranking purposes and tracking. They need to get on this ASAP.

*Moderator Edit: advertising for other gaming platforms is not permitted on BGA.

I was responding to and quoted Den_Man, so I guess you are speaking for them? I wrote a fairly short response and nowhere in it did I say that I disagreed that ELO is meaningless. Even though I think that is an exaggeration, I agree with the spirit of the sentiment. So your lengthy response explaining why it was meaningless didn't actually address my issue.

I can mostly agree that ELO is meaningless and still question why someone would care about that metric if they hold that belief. For me, because ELO has questionable utility I don't make any decisions based on it, which is how I'd expect someone who truly thinks it is meaningless to behave. If someone is behaving differently, then I am curious about why that is.