Player Ratings

[mrg1]

Anybody happen to know if the player ratings system is going to be the

same this year as in 2004? In other words, can I take players from the

2004 game and make copies of them in 2005 without any real difficulties?

I hate when sports games change this. I remember one year ('99 i think)

EA's NHL game went from using numbers to stupid bars to show ratings.

[LetsGoOakland]

I don't think we will know for sure until the game comes out. However, if I hear anything, I will let you know.

[ZMan]

Speaking of this, I am interested in how the roster makers in the community translate their projected player stats into MVP's rating system. Does anybody have any formulas for doing so that they are willing to share?

I have culled together a 2005 player projections database using several sabremetric sources, and would be interested in pulling together a roster based off of this.

However, EA's system is completely Greek to me.

[Sean O]

It's the same system as '04, you can see it in the chyrons that show up as a player gets up to bat in some of the videos.

Z-

For one, that sounds very exciting to have someone so knowledgeable about the game join in the mod community. I would suggest, if there is no direct mathematical correlation, attempting to scale the power ratings by the slugging and the contact ratings by the number of hits or the BA, compared to how the game already factors such things. So if player A in the game has a power rating of 85 and in 2004 real life he hit 26 home runs, you can mathematically scale it from that.

If that makes any sense.

[five]

Does anybody have any formulas for doing so that they are willing to share?

You should PM, them directly. GForce22 probably had the most extensive rosters, but you might also PM SwinginSoriano, I know for a fact that he has shared his formulas with everyone.

I did a quick search for them, but couldn't find the thread I was thinking of.

[paulw]

First of all, I work in Institutional Resarch at a University so I'm nuts about numbers in the first place. I've come up with a formula for power (it's a 4th order polynomial) that I really have been having good results with. Here it is:

POWER = (19+469*ISO-1394*ISO^2+1801*ISO^3)-(2*(5-HR%))

where ISO=Slugging Average - Batting Average

HR% = HR / AB * 100

100 actually means something in the scale. Roger Maris 1961.

The biggest bug is that the minimum is 9. I'd like to set this to zero but would need to change my slopes and/or intercepts.

[mrg1]

cool idea paul

my only question: does that formula give an appropriate distribution? i.e,

is the mean greater than the median, good spread of values, etc....

[paulw]

I haven't run a sample on it. I know the range of values is good but not sure about shape. My guess is a negative skew. I'll try and dig up some parameters for you.

[paulw]

Ran a quick test on the '53 dodgers. It's fairly normally distributed with a mean of 50. Slight negative skew.

[stecropper]

I think MVPEdit gives a fairly representative Power Rating to players with good distribution. The Power Rating assigned to Players via MVPEdit is as follows:

((.SLG% * 1000) - 300) / 5) + 30.5 = POWER Rating

Slugging % > .650 = Rating of 15

Slugging % < .150 = Rating of 0

Hope this helps.

Don

New Orleans

[Cris987]

I think MVPEdit gives a fairly representative Power Rating to players with good distribution. The Power Rating assigned to Players via MVPEdit is as follows:

((.SLG% * 1000) - 300) / 5) + 30.5 = POWER Rating

Slugging % > .650 = Rating of 15

Slugging % < .150 = Rating of 0

Hope this helps.

Don

New Orleans

I dun get what you mean by the "sluggin percentage > .650 = rating 0f 15" stuff, but the former equation kinda...sucks.

According to that, Eric Chavez and Lance Berkman's performances last year only earned them a 73 power, kinda sad, eh?

[stecropper]

Cris987 --- what were their Slugging %'s last year? If their slugging %'s were around .495 then I would think that to be a fairly accurate Power Rating in MVP 2004. A slugging % of around .495 is fairly good but not a super type year.

[Jabroni]

I dun get what you mean by the "sluggin percentage > .650 = rating 0f 15" stuff, but the former equation kinda...sucks.

According to that, Eric Chavez and Lance Berkman's performances last year only earned them a 73 power, kinda sad, eh?

According to this equation, my boy Aaron Rowand of the White Sox will have a 79 power rating:

Aaron Rowand's SLG % was .544 last season so:

((.544 * 1000) - 300) / 5) + 30.5 = 79.3 or 79

You must have screwed up your math with Chavez and Berkman...

Eric Chavez's SLG % was .501 last season so:

((.501 * 1000) - 300) / 5) + 30.5 = 70.7 or 71 if rounded up

Lance Berkman's SLG % was .566 last season so:

((.566 * 1000) - 300) / 5) + 30.5 = 83.7 or 84 if rounded up

I would hope that EA adjusts Chavez's power rating. 71 is too low for him. Chavez's power rating should at least be in the high 70's. Him being injured for half the season likely had alot to do with his low Slugging % last season.

[stecropper]

Jabroni,

Thanks for that info. It makes sense to me. It kinda boils down like this:

Power 90-100 = Slug% .600 - .650

Power 80-89 = Slug% .550 - .600

Power 70-79 = Slug% .500 - .500

Power 60-69 = Slug% .450 - .500

Power 50-59 = Slug% .400 - .450

Power 40-49 = Slug% .350 - .400

Power 30-39 = Slug% .300 - .350

Power 20-29 = Slug% .250 - .300

Power 10-19 = Slug% .200 - .250

Power 00-10 = Slug% .150 - 200

Knowing some of the Oldtime Players like Mays; Mantle; McCovey etc OFTEN had Slugging % Season's > .600 and even into the 700s occassionally would not be out of line for a Super Year to be in the '90s - Exceptional Years to be in the '80s; Very Good Years to be in the '70s; Good Years to be in the 60s; Above Average Years to be in the 50s etc etc etc.

Of course this is all personal perspective and I could see where this scale could range up a bit from where it is to give a bit higher Power Ratings to Players but overall I think it is fairly representative.

I am very interested in others opinions on this subject.

[ZMan]

How might some of the other rating categories work out?

[stecropper]

ZMan,

If you give me a couple ratings you are most interested in then I might be able to provide some of that information relative to how they are derived in MVPEdit if that would interest you. I am always interested in others opinions on how MVP 2004 ratings are derived taking into consideration that MVPEdit has a limited amount of statistics to come up with these ratings as it uses the Lahman Data Base available stats to accomplish this. Perhaps with enough input and consensus we can improve these ratings assigned by MVPEdit for the 2005 version. Just a thought not a promise.

Thanks,

Don

[krawhitham]

Great work guys, I asked for these equations like 10 months ago, and you guys came through.

I would love to have all the equations MVPedit uses

I know when importing some players you get rattings higher than 100

Having the equations would save me time on making an All-time roster

[krawhitham]

First of all, I work in Institutional Resarch at a University so I'm nuts about numbers in the first place. I've come up with a formula for power (it's a 4th order polynomial) that I really have been having good results with. Here it is:

POWER = (19+469*ISO-1394*ISO^2+1801*ISO^3)-(2*(5-HR%))

where ISO=Slugging Average - Batting Average

HR% = HR / AB * 100

100 actually means something in the scale. Roger Maris 1961.

The biggest bug is that the minimum is 9. I'd like to set this to zero but would need to change my slopes and/or intercepts.

if Roger Maris 1961 season is 100 what about Bond's 73 HR season.

[krawhitham]

I get 111.872256 for bond's 2001 season and 99.080474351 for Maris's 1961 season does that sound about right?

[stecropper]

krawhitham,

Appreciate your interest and would value your input greatly. However, I am reluctant to give all the MVPEdit formulas since they actually belong solely to Robert Glass. I did help Robert with these formulas but with the understanding that he is the owner. Not that he asked me to not divulge this info but rather I offerred not to. I still have most of them but as Robert and I were working together I tweaked some that I don't think I have the notes on anymore. I feel OK with providing input to direct questions such as "How are Power Ratings" assigned but would feel a bit out of line in providing all of MVPEdit formulas. Perhaps Robert kept all those formulas and updated them as we tweaked them and is willing to share them. Initially he kept a webpage with them as he & I worked through them but I think that is long gone. I am looking forward to working with Robert again on MVPEdit formulas for MVP 2005 if he is up to it. I have slightly revised some of the Final rating formulas he used with MVPEdit for MVP 2004 and implement those as Global Edits once I import an entire league of Teams into a new .mbe file. So, in my opinion there is room for improvement thus my sincere interest in others opinions.

[stecropper]

krawhitham, I think that sounds about perfect to me. Didn't you let me know long ago that contact rating could go above 100 and if I recall right it was like up to 111. So Bonds at 111 would be correct and Maris at 99 would also be fairly on target in my opinion because there were several other players that had higher Slug %'s in certain years than Maris did in '61 when he hit 61 homers but batted around .261 if I recall. Looks Real Good to me !!!

[krawhitham]

I think MVPEdit gives a fairly representative Power Rating to players with good distribution. The Power Rating assigned to Players via MVPEdit is as follows:

((.SLG% * 1000) - 300) / 5) + 30.5 = POWER Rating

Slugging % > .650 = Rating of 15

Slugging % < .150 = Rating of 0

Hope this helps.

Don

New Orleans

I get 152.5 for bonds VS righthanders in 2001 but when importing his 2001 team with MVPedit the results are 102

his SLG% VS RH was .910

(((.910 * 1000) - 300) / 5) + 30.5 =

((910)-300) / 5) +30.5

(610 / 5) + 30.5

122 + 30.5

152.5

[krawhitham]

krawhitham, I think that sounds about perfect to me. Didn't you let me know long ago that contact rating could go above 100 and if I recall right it was like up to 111. So Bonds at 111 would be correct and Maris at 99 would also be fairly on target in my opinion because there were several other players that had higher Slug %'s in certain years than Maris did in '61 when he hit 61 homers but batted around .261 if I recall. Looks Real Good to me !!!

I was using for those #s

POWER = (19+469*ISO-1394*ISO^2+1801*ISO^3)-(2*(5-HR%))

where ISO=Slugging Average - Batting Average

HR% = HR / AB * 100

[Cris987]

I was using for those #s

POWER = (19+469*ISO-1394*ISO^2+1801*ISO^3)-(2*(5-HR%))

where ISO=Slugging Average - Batting Average

HR% = HR / AB * 100

Bonds stats are just too amazing to compare to the others. No matter what equation you use, you end up at 120+ (unless you have albert pujols at power 50...)

[krawhitham]

Well the game allows upto 127

but my point with the 2nd set of #s was bonds using the equations MVPedit uses should be 152.5 but it was only 102 when imported and I was wondering why

I know Big Clu's power season when imported was 107 for power, so 102 is not the max MVPedit allows