[kereru67]

In chess there are Elo ratings, based on results. It's a blunt instrument but it works. The higher rated player is generally the favourite in a one on one game. In bridge though, it is much more difficult to assess a player's strength. Master points systems, it seems to me, are more a matter of playing a lot than how good you are.

[jillybean]

Yes, MasterPoints accumulate over a lifetime, like the odometer in my car.

Some organizations award MP's and also ranking points. These are based on recent success and will increase and decrease accordingly.
The ACBL has only MP. The players at the top end of the ACBL MP listings have both been playing for a long time and are very good.

[johnu]

OKBridge has a rating system based on ELO for players on that platform. It can be manipulated fairly easily. I had a friend who only played with longtime partners, who tried to play only against pickup pairs of less than expert strength. He was able to get his rating up to top expert levels, even though he was just a good club player, but not considered an expert.

There's an English Bridge Union grading system for their players that attempts to rate players based on recent performances. There was also a website that scraped BBO hands from their database that tried to rate players on BBO. BBO is adamant about not having a rating system, and they eventually barred that site from having access to their hands. That site has disappeared.

In the ACBL, there's something called the Colorado Springs Power Ratings, which used ACBL tournament results to calculate individual and pair ratings. They haven't been updated in a couple of years because the ACBL stopped sharing data with them.

[mycroft]

Yes. What have you won? And with who?

Because frankly, that's the only way to tell.

Some of it is that Bridge Is a Partnership Game, and it's really hard to measure individual skill in a team sport (WAR, anyone? Anyone willing to state for certain that the +3.5 WAR 2B is better than the +3.4 one? or the 1.025 slugging is better than the 1.015?) Some of it is really, truly, because players, by and large, Don't Want to Know (and *certainly* don't want *you* to know).

And, as I always ask when this comes up, why do we care? Note, this is a very serious question. In chess, or shogi, or go, or any of those other games, ELO or equivalent is very useful in making good matches - you're not for the shark tank all day, conversely you're not going to sleep through the event. In bridge, if what you're looking for is a good match for you and your partner, well, that's a laudable goal, and we should be able to do a good job at that. Of course, we don't *do* that IRL - you play 2-3 boards against 8-13 pairs in a session, or you play a set number of boards against everyone in the event (qualifying to long KO matches, usually). It would be a good thing to make work in the Competitive Lounge of BBO (and we've been talking about the "hidden rating system" that BBO could be using to set games - that will only work if they don't publicise it - for 20 years).

Or we could be rating players like the tennis tour, either for seeding or qualification rights or just for bragging rights and to play "is my player going up or down".

But that's not what 90% of players who want a rating system want. What they want is to get a partner "good enough" for them. Maybe even "at their level" (read, "slightly better than I actually am"). And they want to play against opponents "at their level", too (read, "slightly worse than we are, so that we can win"). You can see where this is going. Add to that all the "can't play" issues (technical, like that other thread, or skill-based, like Bridge Master hands, or even "incapable of directing a defence, or reading partner's direction, without using a bullhorn"), and it gets worse (because I may be a good player, but I can't play Polish Club, and I can't play Dutch Doubleton, and I can't understand Vinje signals or odd-even discards (unless the opponents are playing them). And most north americans can't play Keri, or Acol, or any of that other stuff either. And vice versa, frankly).

And then there's the biggest issue with partners, pickup or otherwise (also evident in that other thread). And it can be from the lowest level to the highest - I send you to the Meckstroth vs Barry Crane story (frankly, the entire page of stories).

But finally finally - the fish rule applies. Just like at the poker table, if you can't figure out who the fish is in 4 boards - it's you. Now, you might be stuck with partner (or she with you) for another 22, but oh well...

[mike777]

I have mentioned this rating scale often.

If you make it to the second day of a three day national event AND people are not surprised you are a good player.

If you make it to the third day AND people are not surprised you are an expert level player

Win an open National event and people are not surprised you are a national rated player.

It's not complicated and works well if not PERFECT..

Everyone else is an intermediate or lower level player

Most of us are at this level..


Don't worry about a perfect rating system..

[mikeh]

mike777, on 2026-February-17, 21:44, said:

I have mentioned this rating scale often.

If you make it to the second day of a three day national event AND people are not surprised you are a good player.

If you make it to the third day AND people are not surprised you are an expert level player

Win an open National event and people are not surprised you are a national rated player.

It's not complicated and works well if not PERFECT..

Everyone else is an intermediate or lower level player

Most of us are at this level..


Don't worry about a perfect rating system..

One amendment. Win a National team game and either you’re a national rated player or you’re wealthy enough to hire 5 who are. Believe me, lots of major events have been won by players who are far from expert. Indeed, world championships have been won by non experts, although the ACBL writes about them as if they were.

The same is even more true at regionals.

Don’t get me wrong….in terms of how the game is played at the highest level, professionalism has given many gifted players the ability to play far more bridge than they would ever have played without paying clients, and in turn has allowed them to hone their skills.

[mikeh]

johnu, on 2026-February-17, 20:07, said:

OKBridge has a rating system based on ELO for players on that platform. It can be manipulated fairly easily. I had a friend who only played with longtime partners, who tried to play only against pickup pairs of less than expert strength. He was able to get his rating up to top expert levels, even though he was just a good club player, but not considered an expert.

There's an English Bridge Union grading system for their players that attempts to rate players based on recent performances. There was also a website that scraped BBO hands from their database that tried to rate players on BBO. BBO is adamant about not having a rating system, and they eventually barred that site from having access to their hands. That site has disappeared.

In the ACBL, there's something called the Colorado Springs Power Ratings, which used ACBL tournament results to calculate individual and pair ratings. They haven't been updated in a couple of years because the ACBL stopped sharing data with them.

25+ years ago, before BBO my then partner and I wanted to practice online…we lived in different cities. So we joined OK Bridge, hoping to find good opponents. We soon found real difficulty finding games against good players. As newbies, our Lehman scores were 50, and eventually one opp told us why he wouldn’t play us and why we would find similar reactions from others. Play someone with a lower Lehman and lose, your rating goes down. Beat them, and your rating won’t go up and might even go down a little. Since we’d done well the last time we played this guy….and back then we were pretty good….he saw no reason to play us again.

So we quit the site.

Plus the cheating was astounding!

[pescetom]

mikeh, on 2026-February-17, 23:19, said:

25+ years ago, before BBO my then partner and I wanted to practice online…we lived in different cities. So we joined OK Bridge, hoping to find good opponents. We soon found real difficulty finding games against good players. As newbies, our Lehman scores were 50, and eventually one opp told us why he wouldn’t play us and why we would find similar reactions from others. Play someone with a lower Lehman and lose, your rating goes down. Beat them, and your rating won’t go up and might even go down a little. Since we’d done well the last time we played this guy….and back then we were pretty good….he saw no reason to play us again.

So we quit the site.

But that is a flawed ranking scheme if your ranking cannot initially rise quickly when you do well against highly ranked opponents: it can probably also be tricked by dropping out when you are losing and the other obvious dirty tricks.

The EBU NGS scheme looks meaningful and mathematically valid to me, it would be difficult to extend to all competitive formats but it's fine when you have a series of comparable pairs tournaments with some fluidity in pairing and enough time for it to stabilise (like VACBL on BBO, I imagine).
I wish the FIGB would adopt it for our national simultaneous tournaments, it would be very easy as the EBU has done all the hard work and we have years of data to get off to a fast start.

But they won't, and the reasons are those listed by mycroft: not many people really want to know their true level and even less want others to know it Plus, the RA has a vested interest in keeping people locked into buying MP or national equivalent. To be fair, FIGB does at least track both Master Points and Performance Points, which you only earn by winning important events and never expire. Guess how many players have any PP.

[barmar]

pescetom, on 2026-February-18, 06:54, said:

But that is a flawed ranking scheme if your ranking cannot initially rise quickly when you do well against highly ranked opponents: it can probably also be tricked by dropping out when you are losing and the other obvious dirty tricks.

The problem isn't the ranking system, but the culture that sprung up around it.

No rating system will work well if you're allowed to select who you play against, especially if you can base your decision on their rating.

[pescetom]

barmar, on 2026-February-18, 13:19, said:

The problem isn't the ranking system, but the culture that sprung up around it.

No rating system will work well if you're allowed to select who you play against, especially if you can base your decision on their rating.

Having seen in another card game how easily a similar primitive ranking system can be gamed, I do think OK bridge had a problem of system, not just the related culture.

As for selection, the NGS system is based upon comparable pairs tournaments where you can only select your partner, but I don't see that it would be weakened if you could pick the field (which to some extent you can, by deciding where to play).

Not arguing that it could be applied to all BBO play, but perhaps to many pairs tournaments and at least to the repetitive RA pairs formats like VACB.
I'm not a fan of play against three robots, but again that could easily be compared and ranked on a separate scale.

[johnu]

mikeh, on 2026-February-17, 23:19, said:

Play someone with a lower Lehman and lose, your rating goes down. Beat them, and your rating won’t go up and might even go down a little.

Obviously top experts could lose to lesser players over a few hands, or even over a session or two. In the long term, experts should demolish lesser competition and have their ratings go way up. My friend must have figured out a way to game the Lehman system, but I know he scrupulously avoided playing against pairs of experts, and/or experienced partnerships.
AFAIK, if 4 world class experts joined and only played among themselves, they would all have about average ratings because they are scoring around average against the field

mikeh, on 2026-February-17, 23:19, said:

Plus the cheating was astounding!

I'm not really current with the BBO main room, but there used to be a long stream of complaints and threads about cheaters on BBO, which I found somewhat amusing since there were no BBO points or rating points at stake in those games. Sometimes not even spectators to impress. Maybe BBO has tossed the cheaters out often enough that they aren't coming back under new usernames.
Also, there's plenty of abusive behavior that happens on BBO in casual games where absolutely nothing is at stake.

[mycroft]

The issue with Mike/OKB/Lehmans is that "new handicap" players are a crapshoot. Many of them are okay randoms who will eventually settle to 45, 46, 48 Lehmans. Many of them are good, pleasant players who will eventually settle to 52, 53, 55 Lehmans. And then there are the outliers. The Life Novices who will spend all their time dropping to their (deserved) 40 or less, then every 6 months or so request a reset (which will "new handicap" them again) - and the very good, but new to the platform, players, who will settle into 59+/-, *but while they're doing that, they're pitching ELO-drops left and right to random pairs, because the system "treats them as" the level they "new handicap" to.

So the good players; the ones MikeH and partner at the time need to practise against to actually get better, frankly even the ones that they have to play against to enjoy doing anything but boxing against a blindfolded opponent - they are going to take a huge hit to their ranking from this, clearly within 4 boards good pair, to allow them to climb up to their level. When they could bow out (or not play to start) and let someone else take the hit until they find their level, and *then* play the 59-57 match where a 50-ish score won't hurt.

And the thing is that many of the 50%ers weren't "average OKBridge players", they were new (or reset). The good players don't want to play against 45-47 players (and the 47 pair doesn't either, even though they only have to score 40% to rank up), not just because they have to bunny-bash a 60% against them to avoid losing rating, but because most people feel a little icky about it (and frankly, it's not fun. If you're trying to win a regional open pairs with pros in the field who get paid to bunny-bash, then you play the game you're given. If you're kicking some OKB for an hour or two in the evening with your partner - or if you're practising for the event with expert competition, for that matter - who would select into it?)

Even more so, then, who would select into the 50%ers? On average, they're 45-47s that haven't reached their level (again) yet. Some are worse, but after one "hopeless" level, who cares? And some - some are the Spikes in the weeds that will eat our ratings to feed theirs. Nah, someone else can do that.

Lehman rating is a decent attempt. I should know; I used to host the "how it works" site back in the day. And yes, there was a "rating limiter" for "new handicap" players, where the swing of the opponents' rating would be cushioned, and the new players' rating would swing more wildly, until they had enough boards to "confident up" the score. Which clearly didn't help, because [mikeh's story, among many others].

My stories of course about rated games on OKB also apply, again mostly due to inability to actually *understand* the rating system:

• Nobody would play with me (54%) and my newer player girlfriend (46%). The rating bands requested were *tight*, and either "you're too good" (despite my partner's rating) or "we want a good game" (despite mine).
  
• Of course, with the bands so tight, people basically had even games against each other all the time and the ratings never changed. Which, fine, but the whole crew of these people never found out if they were getting better...
  
• A game with my mentor, where we talked for a bit after each hand. Not only did we have the rating discrepancy in the previous, but even with the description, they'd say "talk later, play now". And if we set an unrated game, we never got opponents to start with. Eventually we had to team up with another mentor pair, and then when her mentor gave up, I acquired her and her regular partner as well, so it was "me and three".

And yes, because you are what you measure, *any* rating system will be gamed by some people, and frankly exactly how JohnU's story did it - because again, they don't want *you* to know how bad they actually are (see? I'm a 61 Lehman player!) Of course, if fortune (or fame, but usually fortune) were on the line, the same people would submarine their ratings to enter events they could clean up in - just like they do in golf.

So, we're back to: "What have you won?" and "with who?" Which works for the Korbels and Brogelands of the world. Frankly, it's how they make their living - "what have you won" and "who will/won't play on your team" are precisely the answers to key questions for sponsors when deciding who to hire.

It doesn't work for the kererus of the world, who want to know if mycroft is good enough that the next half hour to two hours won't be a nightmare, or for the mycrofts of the world, who want to know if kereru is going to point out every mistake (he thinks) I make for the next two hours, or is slower than molasses in Iqaluit (whether he's the best player in the room or, usually, not), or is even going to be the subject of Helene's next rant while I get tarred with "is willing to play with this pest". None of which I have any reason to believe OP is like, please note!

[barmar]

johnu, on 2026-February-18, 18:29, said:

Obviously top experts could lose to lesser players over a few hands, or even over a session or two. In the long term, experts should demolish lesser competition and have their ratings go way up.

Your rating doesn't go up much, if at all, when you beat someone you're expected to beat. Your rating goes up when you do better than expected, and it goes down when you do worse. Getting the expected result implies that the ratings accurately describe both pairs, so they stay the same.

So the Lehman system discourages playing against people much below your rating -- there's no upside if you beat them, only downside if they happen to do well. They might be expert newbies whose ratings don't yet reflect their expertise, and they'll take you down while they go up.

I don't remember how Lehman handles big discrepancies between members of a partnership.

[mike777]

I still remain convinced that while my rating system is not perfect, it is a heck better than any other ideas.
And much simpler.. smile

[barmar]

ACBL supposedly has talked about adding a rating system for a while now. The editorial in this month's Bulletin says they're working on it again, apparently based on a proposal by Nick Straguzzi. No details are provided, just an assurance that it will be in addition to Masterpoints, not a replacement.

[pescetom]

barmar, on 2026-February-19, 15:02, said:

ACBL supposedly has talked about adding a rating system for a while now. The editorial in this month's Bulletin says they're working on it again, apparently based on a proposal by Nick Straguzzi. No details are provided, just an assurance that it will be in addition to Masterpoints, not a replacement.

IIRC he outlined his proposals on that other site a year or two ago.
It looked to me much like NGS but with a less candid description of how it works.

I still fail to see why any RA can claim a legitimate reason to not implement NGS or why the more intelligent players (and club leaders) should not desire it.

[AL78]

You can do it approximately because unlike chess, bridge is not a deterministic game. There is considerable variance (maybe less at the expert level) that can impact your results and rating.

An example from my recent situation, pickup partners in a club with a low standard (NGS ~ 44%), my NGS just under 57%, I need to get near 60% every time to break even. Given an evening with a mean hand HCP < 9, declaring three times in 24 boards, cards go the way of the opponents, they bid everything that is available to them, half the room can't bid 27 HCP games, good luck with getting over 50%. My NGS has dropped over 1% over two sessions largely thanks to stupid annoying sessions like this being on the wrong end of hand biases where it is very difficult to near impossible to engineer anything on many hands. On the other hand, partner an dI get the cards, the pickup partner is competant, a couple of squeeze plays come up, we end up with over 70%. NGS doesn't care about level of variance or whether you are fortunate enough to have an enthusiastic regular partner you can work and grow with.

[jillybean]

I can see Sponsoring Organizations using Performance Points for Stratification purposes. This will also keep players who would prefer to play amongst their peers rather than play up, happy. Other than that, who cares? You can looking at the rankings and see the top players, you can play a game with, or against anyone and quickly determine if they are a candidate for a serious partnership, or pleasant and decent enough to play with on a casual basis.

[mike777]

Is this really a priority for bridge?

2025 ACBL lost another net 7% in membership

[mikeh]

pescetom, on 2026-February-19, 16:01, said:

IIRC he outlined his proposals on that other site a year or two ago.
It looked to me much like NGS but with a less candid description of how it works.

I still fail to see why any RA can claim a legitimate reason to not implement NGS or why the more intelligent players (and club leaders) should not desire it.

I have zero interest in a ranking method in ACBL land. I fail to see what useful purpose it would serve other than creating more opportunities than currently exist for arguments.

Good players recognize other good players. If I played 6 hands against you and your regular partner, with me playing with mine, I’d almost certainly have a reasonable idea of your skill before we finished….and I’d expect that to be more accurate than whatever ranking you have.

Meanwhile, rich client A, who would struggle to break average in a Sectional Open Pairs playing with a peer, wins 100+ masterpoints and two or three team games at a regional and the ranking system rates him the same as the five world champions he or she paid to win those events. As someone who, at least in local Regionals, is always in the top bracket of a KO, I have played against many pro teams. I know of which I speak.

In my last local regional, there were 6 team events. My amateur team* won 3 and a client won the other three. That client wouldn’t have been accepted on our team, absent payment, yet any ranking system that the ACBL might implement would treat him as one of the strongest players in NA…he travels to a LOT of events and always has strong pros on his team.

Why would any intelligent bridge player want such a system?


Btw, I’m not knocking that client. He seems like a good guy and most smart people, even back when bridge was far more popular than it is today, spent their time doing things other than playing bridge. It’s just that if they’re not really good, don’t make a ranking system that ranks them as being really, really good. We already have that with masterpoints.

* 2 players on the team do play a little pro but they don’t make a living at it nor were they getting paid then.