Raw player rating formula for hitters and pitchers

[DALTONPARK]

I have purchased the ratings disk and am looking to evaluate players. Is there a straight formula used for ranking player cards vs righty only and vs lefty only. For hitters and pitchers.

Example What hitter if you could bat him in all 9 slots would you want to face a right handed pitcher. What formula is used to figure this out. Also what pitcher facing a line up of all right handed hitters would you want. And what formula is used to decide this.

I have read Deans post on Erp and Nerp but he ranks players based on total season facing both left and right handed pitchers. I want to look at it like a platoon situation. Where they rank against lefties only and where they rank against righties only. Also i wouldn't mind seeing defensive runs given up subtracted from these ratings.

Does anyone have anything like this? Could you post it if you do? Or help get me started in the right direction.

Thanks

[maligned]

If you understand how Dean is using NERP to evaluate a whole card, it should be very straightforward for evaluating half of a card. Just take the NERP value of "vs. right" for all players to see who has the strongest offensive value against righties and the NERP value for "vs. left" to compare against lefties. Then, use his NERP suggestion for defensive value and subtract it from your offensive total against a given side to see who has the best overall value at a given position against righties or lefties.
On your spreadsheet, simply create a column that calculates offensive value of the "vs. righty" variables only. Do the same for "vs. lefty" values. Then you can sort them as you like to see who is the best against a particular hand.

[tcochran]

The original link for the NERP article leads to an abandoned domain. Is there a new link available?

[DALTONPARK]

1st Tcochran there is a post on down in this forum called offense verses defense updated with a live link.

malighned Would I use 648 for the number of AB's or 108 for card chances in the AB spot of the formula? I know Dean used 648 for hitting purposes. Does that match what he used for defensive ratings.

[tcochran]

In the offense vs. defense article, other links work but not Dean's -- that's the same dead link.

Anywhere else?

[DALTONPARK]

http://silenceisdefeat.org/~meandean/som/OffensevsDefenseinSOM.zip

[DALTONPARK]

My article on how to evaluate the offensive and defensive contributions of a SOM card has been updated and greatly revised. The article can be found at:

http://silenceisdefeat.com/~meandean/som/OffensevsDefense.zip

[tcochran]

Outstanding! Thank you...

[gbrookes]

I also want to say a very big THANKS (!!!!), and provide some feedback:

-I have been using my own mathematically simplified version of these formulas (still largely true to their original math, I believe) for about 9 months, in my strat playing and team managing. I actually often do the calculations for the 3 games being played on a given night. I find that the formula accurately predicts the outcome of the game reasonably consistently, taking into consideration the usual luck factors - i.e. if it accurately predicts the outcome 60% of the time, it is pretty darn good. It does for me! Better still, it really does a good job of predicting the flavour and style of a game.

-Of course, you can predict the flavour and style (and maybe the outcome pretty darn well) without the formula, just by looking at the cards. What this formula does so well from my perspective, though, is to make it an OBJECTIVE, QUANTITATIVE and RELATIVE analysis. The numbers I calculate often times directly influence my plans, including lineup changes, pitcher settings, etc. I feel very strongly that this has helped me to make those decisions better, and help win games from time to time.

-I think it is important to read the comments in the paper about what things were NOT considered. My own view is that, with the extra running rules used in online strat (just about any hit can result in an extra base running opportunity for runners on base), baserunning ability and outfielder's arms are more important than are implied by the comments in the paper. In any event, he acknowledges that it would be an interesting subject for further study and evaluation. For the past year, I have been weighting my player selections to highly value good base running and good outfielder arms.

-Overall - love it. Great tool for planning teams and even games in strat, in my view.

BTW, my simplified formula (applied to results times die roll weighting on player's cards, with ballpark effects included) is:

Relative offensive contribution = walks(&HBP) * 1.2 + singles *2 +doubles *3.6 + triples * 5.4 + homeruns * 7.

I know this formula doesn't do any justice to the NERP in the paper posted above, but it works well enough for me in planning my teams and my games.

For other more complex effects, including GB DPs, I just use my instincts to evaluate this, and of course try to plan around it with lineup order, bunts, hit and runs, etc.

As a few happy coincidences:

-this typically works out to about 100 for an average to good player. (as a matter of convenience for myself, I include ballpark singles for both the pitchers as well as the hitters cards in the batter's card evaluation total. Then I do a separate adjustment for any situations where there are no ballpark singles on either the batter or pitcher cards. That way, the standard addition for ballpark singles is ONLY on the batter card, and it is always the same for each lefty hitter and for each righty hitter.

- If a hitter is W power, I just subtract 10 from his total to take this into account, as a rough estimate of a typical effect. In game situations, i calculate the pitcher's card value against either W or N power hitters, and then use the actual pitcher card value for that hitter.

-VERY INTERSTING - I find that this simplified formula does a VERY, VERY good job of estimating the L/R balance for hitters. Take the batters card evaluation (as per above) for lefty hitters, and subtract that from his evaluation for righty hitters. Take the difference and divide by 10, rounding to the nearest whole number. This result is the SAME as the strat L/R rating system for that batter, about 70% of the time!!!!!

-The above comment works the same for pitcher cards, just as well, but instead of dividing by 10, divide by 7. The L/R balance works about 70% of the time.


Geoff

[gbrookes]

By the way, I also adopted my own simplified analysis of fielding effects, although I certainly make use of the X chart excel spreadsheet that you can access from the above link. I just basically do my own simple math using the simplified formula I posted above.

I also want to acknoweldge that I am NOT creating anything NEW with my own simplified version of the formula.

I also want to disclaim the accuracy of my own version of the formula, compared to the one posted in the paper. My version may not be accurate, although I think it is mostly accurate. I know that there were arbitrary assumptions that I had to make in order to convert and simplify the formula.

I am only sharing this because I found something that was simple enough to use on a nightly basis, and for the sake of building my teams and evaluating trades and player personnel questions.

Cheers and good luck!!!

Geoff