Something Semi-Interesing-ish

Still not perfect, of course. Ideally, we would include hands where we bet, but everyone folds. I guess we have to work with what we have though. For the use I have in mind, it probably doesn’t make much difference.

Thanks for the constructive input!

IMO, for tournament players who would like to track their progress using an easy and already available method, tournament points are the best way to do it.

Tourney points are calculated while taking into consideration the buy-in, the position one finishes in and the number of players in that tourney. All these factors are important in determining how well one does in a tournament, and they’re already available.

All one needs to do is to see how many points they are accumulating each month, and compare it with the previous month to track their progress.

For those who are crazy about averages, you could always divide the total number of points by the total number of tourneys played in a specific week or month, and you will get the average number of points you are making per tourney. That could be a useful tool to compare and track.

Chips won in tournaments are prizes, and hands played in tournaments don’t generate anything towards the bank, so it’s irrelevant to create complicated ways to relate them, regardless of how interesting they are.

P.S. This message is for serious grown-ups who don’t mind discussing different ideas and solutions on the forums, and are serious about tracking their progress in methods that actually work, like @puggywug and @Fozman Not for the haters who only weigh in to insult people they don’t like, without any relevance to the topic itself :wink:

Tournament points are good for tracking your own progress. And you can track the leaderboards in the weekly promotions, which is very handy.

I don’t think anyone’s brought it up, but my main problem with the Toplists page is that they’re dominated by players who have an ungodly number of chips, and have played for ages. I wish the leaderboard page had some time-limited boards as well, to let us get a sense of who’s hot this week/month/year.

They’re less useful for gauging any two random players. A player who has a long history will have had more time to accumulate points than a new player, etc. Or a player who just doesn’t play a lot of tournaments won’t have as many relative to a player who exclusively plays them. Etc.

Plus, I’m not sure that there’s a convenient way to see another player’s tournament points. I guess there’s not a convenient way to get the other information that SPG uses for his calculations, either, though.

I do think the site would do well to implement some new system for establishing player ratings, beyond tournament points, the leaderboards, and the bankroll.

1 Like

That’s true! But you don’t need to be on any leaderboard to track your own progress using tournament points.
You could just make note of points earned in every tourney you play, like on an Excel sheet, and track your own progress without necessarily having played enough to be on any leaderboard.

You’re right, those boards are mostly topped by those who play a lot, and there’s no way to see another player’s points if they’re not on the board, but for self evaluation and tracking own progress, tourney points vs number of tourneys played could still be a useful tool that doesn’t require massive or complicated calculations.

It’s still only for tourneys though. But Replay is introducing Ring Game Points as well, which could be equally useful, and could save everyone from having to go back to their banks and hands and create their own methods.

1 Like

I’m only interested in tracking my own progress. Tournament points are sensitive to the number of tournies played, profit or loss per hand isn’t.

The MTT monthly boards will only count a specific number of tournies. The SnG monthly boards will show averages over all SnGs played in that bracket, but it’s still hard to translate tournament points into meaningful data. One could show a gain in average tournament points and still not know if they are winning or losing.

Dividing chips won by number of hands played is simple, requires minimal record keeping, and provides a meaningful metric.

It’s not dividing them that’s the problem, it’s getting those accurate numbers.

To get the total chips won, you need to go through your history, and take out all the freebies and rewards etc…

Still incomplete and inaccurate.

To get the total of hands played, you either have to get the number of flops seen, which is inaccurate because it doesn’t include the number of hands you won by making everyone fold, or get the number of hands played which also doesn’t include all the hands that have seen the flop.

So far, there’s nothing simple about this or that requires minimal record keeping.

Now to go back to repeating myself and stating the obvious facts which unfortunately keep being ignored for some reason, in tournaments, chips won are not won per hand, they are rewards for making the top paid places. They are called TOP PRIZES for a reason. They are just prizes. They are completely irrelevant to the number of hands played.

Dividing chips won in a tournament by the number of hands played in a tournament to track progress is absurd. The number obtained is as relevant to our progress as dividing any 2 numbers that bear no relevance to each other and using the result.

If I play 1 tourney this month where I have played 60 hands and didn’t make it to the top paid prizes, that gives me 0 chips won per hand, even if I have won 40 of these 60 hands, still no chips in the bank.

If next month I play 1 tourney and I win first prize of 1M chips in just 4 hands, that should mean that I made 250k chips per hand.

What a great way to measure my progress!!! from 0 chips per hand to 250k per hand.

Again, the chips won in a tournament are NOT won per hand. They are prizes. They cannot be divided by the number of hands played. It’s totally absurd and irrelevant.

But hey, what do I know? Let’s keep pretending these aren’t the facts, and keep insisting that the number is in any way relevant.

1 Like

Having been gainfully employed for several years tasked with dissecting and manipulating mountains of data into meaningful results used to drive strategic decisions, I guess I could be considered a “professional” in such affairs. I’ll share some of my thought process with you here.

Clearly, it’s obvious the metric discussed so far in this thread is only going to be useful for Ring Games. As such, we need to focus on the chips earned only in ring games and some statistic to divide by. Fortunately, I’ve started the work on that already earlier in the thread.

Something else to consider: the stake levels one plays at are going to heavily skew a raw chips earned metric. We would need to make this figure “dimensionless”, if you will, by exchanging this for something used across all ring games: the Big Blind.

At this point, we are dabbling near the BB won & Ring Game Points metrics that have been used on Leaderboards. We want to arrive at BB won per played hand.

This section is mostly moot because of the following paragraph... click to expand.

So, we must decide what to divide by:

  • Total Hands – moderately meaningful and all-inclusive
  • Non-Folded aka “Played” Hands – not a very accurate estimate of actual played hands because many folded hands are also “played” and have a chip cost associated with them, however if one could derive an accurate count of hands in which one is involved this would be the best metric
  • Flops Seen – is a relatively good estimate on “pay to play” hands, but has some degree of inaccuracy because of what can happen pre-flop, despite this it could be a pretty good metric
  • Pots Won – does not take into account the majority of hands played, however it still could be a marginally meaningful metric

Let’s also look at some logical definitions:

  • Folded hands are always losing hands and sometimes change the chip count.
  • Pots won are always non-folded hands and always change the chip count.
  • Non-folded hands contain both winning hands and losing hands and always result in a change of chip count.
  • Flopped hands contain winning hands and losing hands and always result in a change of chip count.

If we consider any hand that changes the chip count to be a “played” hand, the problem we run into is that some of these hands are “stuck” in the folded hands statistic. Besides that, we cannot use the statistics page because it also includes tournament hands. Trying to generate an exact count is impossible from the statistics page. We could try using another derived statistic or try to create a ratio of winning hands vs. losing hands to multiply by, but these aren’t as meaningful.

Enter the short term. We can get an accurate count of “played” hands in our recent ring game activity page; simply discard any hand with zero net chips. We can also get an accurate count of chips won and thus calculate BB won. It’s relatively easy to do manually on a single session or a few short ones. (Automating it is incredibly difficult due to formatting, but I already have a way to extract all the values and suits of the cards… so probably not a huge leap.)

Here’s an example of a short term calculation:

Other than a slight distortion from folded blinds, this seems like the most accurate way to gauge ring game effectiveness. Aggregate enough of these sessions and you will have a long-term progress report that you can plot out over time or even see the tables at which you are most effective.

3 Likes

First of all Fozman, thanks! Both of your posts took a lot of time and effort, and I, for one, appreciate it.

After thinking about it more, I am going to just use “total hands” and be done with it. I don’t have the desire to spend much time on this, or to do any more work than I absolutely have to. All I want is a down and dirty metric that I can track over time to see how my adjustments are panning out. (or not) As long as it’s fast, easy, and delivers a consistent result, it will meet my needs.

I tend to specialize. Right now, my focus is on 25K 9 seat SnGs. Since that’s all i have been playing the last few weeks, a lot of the hard stuff is automatically eliminated.

This is neither clear, nor obvious to me. Total chips won divided by total hands played meets my needs nicely. Would you care to elaborate on why you think this is a bad approach?

It would be great if the site allowed people to track their results for rings and for MTT’s/SnGs. Let people see their BB/100 for rings and ROI’s on tournaments. If I’ve heard correctly, pretty much everything is on hold until they do an upgrade for the whole site. After that is completed, there are many things that would be great to see added. Most are pretty simple because the data is already being collected. They could have a statistics page customizable by the players that could show whatever metrics that player wanted to track.

3 Likes

I put some thought into this and I may have inadvertently proven your point.

I think the initial definition of the problem led me to agreeing with Maya’s objections regarding the “localized” nature of tournament chips and hands played.

This updated definition – total chips per total hands – did make sense once I looked at it. In order for it to make sense, one must define “a hand” to be some unit of time. We can translate this to a standard unit of time if desired – “a hand” is approximately (1) minute – but it isn’t necessary. Basically, what this calculation tells us is: “How much am I earning while I’m actively playing in any game mode?” In other words, this is our “hourly wage”.

Using this definition, we can seek out ways to maximize chips per hand.

  • In ring games: win as much as possible in as few hands as possible at the highest stakes possible.
  • In SnG/MTT: place as high as possible while removing competition as quickly as possible at the highest stakes possible.

The ring game objective is very intuitive. However, the tournament side may not be. Let’s look at the factors that can affect the outcome of “chips per hand” in a single tournament:

  • Blind schedule = faster means effective stacks become smaller leading to quicker KOs and fewer hands (faster is better!)
  • Effective starting stacks = smaller stacks mean quicker KOs and likely fewer hands (smaller is better!)
  • Table sizes = 4-Max & 6-Max mean more tables to play through and quicker hands meaning more hands dealt but potentially more opportunities for KOs per hand tournament wide leading to quicker prize money (break even to slight edge for 9-Max?)
  • Entrants = fewer entrants likely means fewer hands, more entrants likely means larger prizes (break even to slight edge for fewer?)
  • Bounty tournaments provide synergy with maximizing this stat: one earns chips directly for removing competition.

As such, turbo bounty tournaments at the highest stakes possible should maximize the “chips per hand” stat if one focuses solely on tournaments. Basically, you want to get the largest tournament prizes you can in as few hands as possible.

I think this is the core of SPG’s point, Maya. He wants to figure out his “hourly wage” for all game modes and play game modes that net him the most chips in the least amount of hands (i.e. time).

1 Like

I understand that, but it doesn’t work that way, especially for MTTs.

The prize is fixed, you can’t change it, you have no influence on it whatsoever. Top x places are paid fixed prizes at the end of the tournament. Whether you play 5 hands or 500 hands, the prize is the same.

My point is, you can’t divide the number of chips won as a reward in an MTT, by the number of hands played which did NOT generate the chips. You play to place, not to generate chips.

Everything plays a factor in an MTT except the number of hands you play. The number of players, the buy-in, the spot you finish in, they all play a role in determining which prize you win, but the number of hands you play doesn’t, even if you consider it as a “time factor”.

The point of this exercise is to track your progress, and how many hands in takes you to finish playing a tournament is insignificant. I gave many examples. You could play 5 tournaments per month and not win in any of them. Even if you played a total of 500 hands and won 400 out of these 500 hands, if you don’t place, that’s 0 chips per hand. It doesn’t mean you’re playing worse than before. You still won 400 hands. But there are no chips to prove it.

So, while I agree with you that winning a tourney prize in as few hands as possible could in certain cases be an achievement, I still don’t see you have any control over the number of hands you play like you do in ring games, especially when there are multiple tables and you can only affect your own table.

Another difference is that in ring games you can take your wins and leave at any point. In a tournament, you can’t do that. You are stuck and the number of hands you play will be counted no matter what.

You can still do that and not place and not win any chips. How would that be counted toward your progress? If you do all that and you still bubble for example, 0 chips won for all the hands you played, how does that serve the purpose of tracking your progress? Placing 4th in a tourney where only top 3 places are paid should be counted toward your progress even if you don’t collect any prize. But in the method suggested it doesn’t. It’s like you’ve done nothing. You played 50 hands and got 0 chips.

1 Like

If you look at a single data point, yes, it is meaningless. 1 MTT, 2.5hr = 200 hands played, 0 chips won. RG, 2.5hr = 200 hands played, 125,000 chips won. This snapshot overwhelmingly says “play ring games”.

Quite often, the average AND/OR the Dx/Dt is what provides meaning to a statistic.

Let’s say a product of mine has 30 serviceable incidents per million in a single month. By itself, it means nothing. But if I tell you last month, the same product only had 10 serviceable incidents and the month before that only 5 it begins to tell a story. We may have a failure mode that is getting worse. The Dx/Dt is what matters with this case.

Coming back to poker, examining the numbers for a single tournament in which you don’t get prize money is always going to look bad. Let’s say I start a new account and play the Bankroll Builder 1K tournament 20 times. I win prize money a few times and net about 75K total. I also played 7.5K hands in these tournaments. I’m getting 10 chips per hand. This average case is a little more meaningful, but not very much until I have something to compare.

Let’s say I continue by playing some 2/4 ring games for a while. I play about 5K hands and come out ahead by 25K chips. I’m getting 5 chips per hand. At this point I would say to myself, I’m probably earning chips faster over the long haul by playing the 1K tournaments.

Edit: First paragraph.

1 Like

Just a quick example. Let’s say I’m tracking my progress using this method.

First month I play 5 tournaments, 100 hands each, and I place 4th in all of them. They all have top 3 places paid so I play 500 hands and win 0 chips, but I placed 4th in 5 tournaments.

Second month I play 10 tournaments, a total of 500 hands as well, and I place last in 9 tournaments, but I win one of them and the prize is 50K chips. So in that month, using the suggested method, I got 100 chips per hand.

By this methods’ standards, my progress is that I went from 0 chips per hand to 100 chips per hand, which should mean that I’ve played better and I’m improving.

But placing 4th in 5 tournaments on the first month is much better than placing last in 9 tournaments on the second, so that contradicts the progress obtained.

1 Like

Absolutely. We have a metric for that called tournament points. The 4th in 5 is better play, sure, but also less profitable for the effort involved. We are attempting to derive a metric for maximizing chips per amount of time played… be that in a tournament or a ring game. For this new metric, we don’t really care too terribly much about how well we actually play… we just want chips.

We can debate the merits of maximizing chips rather than playing well, but that might be derailing the thread a bit. Good discussion, Maya.

1 Like

Thanks Foz you too :handshake:

Yes, exactly. And Foz, if you start talking about 6 sigma, I may ask you to marry me.

I started this thread with no real goal in mind. I threw something out there, and let’s see where it goes. It’s now gone beyond semi-interesting-ish and is moving towards really interesting-est.

I issued a challenge: win 20 9 seat SnGs in a week. What does it take to do that?

I do have an even simpler metric I can and do use. For example, last week I played 51 25K SnGs and won 15 of them. That’s a 29.4% win rate. Let’s call that my baseline. So 15 isn’t 20, how do i bridge that gap?

One solution would be to play more tournaments, but that can’t scale forever. i don’t think I have more than 60 or 70 a week in me, and I know that, if I play too many, I get real sloppy.

So, this week, I am amping up my aggro level and opening my ranges a little. I have a baseline, so I can know if these adjustments take me in the right direction or not. Yeah, there will be variance no matter what, so the results won’t be bomb proof, but it’s something.

If I don’t get the results I’m looking for, next week I will tighten my ranges and play a little more passively.

In this specific use example, I am using win rate as the metric. It’s fast, simple, easy… all things I like.

Bit I would also like to develop a more general metric. To me, saying “I played X number of hands and won Y amount of chips” seems to fit that need. If I was playing a wider variety of tournament types, buy-ins, and so on, I might need something that would tell me which format is best for me. Since I tend to focus on one format or another, I don’t need that degree of information, and so don’t want to spend the additional time and effort to develop a system that has more buttons than I will press.

However, it’s not always about me. I think this thread has taken a turn that could lead to something of more general interest to a wider range of players. It goes where it will, and that’s fine with me.

So now, the question becomes: what’s the easiest way to track your progress over a variety of formats, and to see which one returns the maximum profit?

1 Like

Net $(chips) / hr. - From what I’ve been reading, this is the way many pros determine what games are and are not worth playing. A pro’s limiting factor is time. He can play rings or tournaments. He can play 1 table or multiple tables. The only thing he can’t do is create more time. So, if he’s playing 4-tables at 500NL and netting $100/hr. regularly, in order for another game to be worthwhile for him, he must expect to make more than that.

I think this works pretty well because you can move up and down in stakes and change game types but boil it all down to an hourly win-rate. If you keep track of which games you have played, the time you have invested and the net chips won, you should be able to get a good picture of which games and stakes you are maximally profitable playing. This might not be very convenient to do here because it would require manually tracking the times spent in each games but if you set up a simple spreadsheet and keep up with the data, I think it would be very useful.

1 Like

Yeah that works if you don’t mind the record-keeping.

I count the chips I get from leaderboards, which makes it a little more complicated anyway, but I guess I could divide that by the number of hours too.

Honestly though, there isn’t a really easy way with the stats they provide.

Now, if they give us the ability to use something like PokerTracker, or at least let us DL hand histories, life would be good!

2 Likes

also, even most sng have leaderboards, so, if one is as good as they say, they’d clearly be getting add chips there as well.

That assumes the player has the time to play the required number of games per week/month.

1 Like