MWC vs Equity

[PersianLord]

I'm used to base my interpretations of the game on MWC (Match Winning Chances) provided by GNUBG when analyzing. But some of my friends have kindly adviced me that using Equity is a way better. I'm wondering why?!

I think that equity numbers and MWC is somehow the same entity, as GNUBG's default setting of analysis provides you with MWC. So, if MWC is based on equity data, why bother to read it instead of the straight-forward and more easy-to-understand MWC?!

Any imput would be great. Please educate me on this issue.

Regards

[spielberg]

I agree with you and prefer MWC to an equity number when looking at point matches. Equity becomes important when you're looking at cash games, I feel. Anyone care to differ?

[PersianLord]

I agree with you and prefer MWC to an equity number when looking at point matches. Equity becomes important when you're looking at cash games, I feel. Anyone care to differ?

Thanks for imput. I think that this article makes it clear that MWC is actually the same with equity:

http://www.backgammoned.net/backgammon-articles/backgammon-equity.html

We read that:

Match equity is often referred to as called match-winning chances or MWC%.

So what's the difference?!

Any other response?

[don]

The way I read the documentation at http://www.gnubg.org/win32/gnubg/gnubg.html#Equities, when gnubg evaluates equity, it is talking money equity, not match equity.  Thus MWC is different from equity for gnu.

At least this is what I infer from the way the doc-gnu-mentation distinguishes between the two.

--
don

[blitzxz]

Gnu uses both match winning chances (MWC) and normalized money game equity (EMG). The difference is that in EMG the score or the cube doesn't affect the equity. EMG is typically considered better because you get better picture of how large error you did. For example if you make a huge blunder in first game of a long match you're not likely to lose more then couple prosents of MWC. But if you make the same error in double match point you can lose 30 or 40 prosent of MWC. This can be confusing and doesn't help at all unless you can decide when you make your errors. MWC is propably better when you're measuring luck and it might be useful when trying to figure out take points in different scores.

[PersianLord]

Gnu uses both match winning chances (MWC) and normalized money game equity (EMG). The difference is that in EMG the score or the cube doesn't affect the equity. EMG is typically considered better because you get better picture of how large error you did. For example if you make a huge blunder in first game of a long match you're not likely to lose more then couple prosents of MWC. But if you make the same error in double match point you can lose 30 or 40 prosent of MWC. This can be confusing and doesn't help at all unless you can decide when you make your errors. MWC is propably better when you're measuring luck and it might be useful when trying to figure out take points in different scores.

ty. But what about GWC? (Game winning chances) I mean that GWC must be the same as EMG as it calculates your chances of winning the current game. So taking into account both GWC and MWC is better, isn't it?

[spielberg]

MWC is NOT the same as equity as don's post clarifies. Equity is a cash value and MWC a percentage. A fine post from blitxzx has prompted me to use EMG in my endless quest to indentify the weaknesses in my game (and those of others hopefully!).

[blitzxz]

ty. But what about GWC? (Game winning chances) I mean that GWC must be the same as EMG as it calculates your chances of winning the current game. So taking into account both GWC and MWC is better, isn't it?

Basically what ever you use it doesn't affect what gnu thinks is the best move. So you can use what suits to you.

And let's make clear what is equity. (Game) Equity means expected number of won points in that game. Normally cube is not counted, so equity can be any number between -3 and 3. (from sure backgammon loss to sure backgammon win). For example is you get doubled and your equity would be less the -1 then you should pass and give the point (Now of course, you have to count the cube). This is different then game winning chances unless you some how count the gammons and backgammons also. Or you're playing to one point match when game equity truly is the same as game winning chances.

And to make it more unclear match winning chances are also sort of equity. It is same as expected number matches won. When playing particular one match it can only be from 0 to 1. For this reason match winning chances are sometimes called match equity.

[PersianLord]

And to make it more unclear match winning chances are also sort of equity. It is same as expected number matches won. When playing particular one match it can only be from 0 to 1. For this reason match winning chances are sometimes called match equity.

Thanks. My point is exactly this. I mean if, as you say, MWC is the same with match equity, so I can base my interpretations on MWC. But many high-rating figures in DailyGammon forum have adviced me to use equity instead. If these are the same, what difference would it make?! 

Or may be they are referring to game equity,  rather than match equity?!

[blitzxz]

Or may be they are referring to game equity,  rather than match equity?!

Yes, they propably are refering to EMG, which is normalized money game equity, which is good indicator of errors also in match play. 

[PersianLord]

Yes, they propably are refering to EMG, which is normalized money game equity, which is good indicator of errors also in match play. 

ty.

[vikingblood80]

To put it simply:

Using EMG makes the size of errors compareable (which is not the case when using MWC)....a -0.1 blunder is a bad move not taking actual match score in account. At different scores this -0.1 blunder can lead to a loss of 3% MWC or 2% MWC or may be even less than 1% MWC cause here value of gammon's, single wins and all that match-score dependent stuff are taken into account. So with MWC you cant tell if your move was a generally bad one or bad depending on score.Finding your errors this may lead you on the wrong path.

The other way round a error of 2% MWC can be a blunder of -0.2 or -0.08 or at DMP may be only -0.04..you simply don't know it.

If you want to spot your conceptional misplays then EMG is better to remove your whoppers from your play. Nevertheless MWC also has its advantage and I sometimes have a look at it for difficult cubes. 

CU Viking