Attributes and Ratings

[ksig24]

Does anyone, specifically player editors or CAP makers have information on how these attribute ratings work?

The Edit Player section is extremely comprehensive, much more than I am used to.

I noticed a lot of 90+ OVR rated guys and most of them have high contact rates. What do these contact rates equate to? Is 99 equivalent to .325 or so?

Also, after a season sim I noticed a major power outage from superstar sluggers! Prince Fielder had 20 homers, Adrian Gonzalez 16 etc...

Out By K - what rating does a 100+ K guy get? Does it play into simulation results?

What rating for a player with 100+ BB?

Stealing - ?

Do players with 4-5 star potential progress?

I am all for opening up for discussion. I would like to know this because I will be buying 2k12 and would like to contribute to the roster making.

[peigone]

Re: Attributes and Ratings

I am all for opening up for discussion.

I've been buying this game since the Blue Shift days and have been editing ratings since 2K5. There are really two conversations one could have here. One can create rosters based on the default, over-the-top ratings system 2K uses which includes too many 85+ rated players, and too many players with high contact/power numbers as you pointed out. The ratings system 2K uses is specifically designed to create accurate stats when simming, even though those results aren't all that accurate as you again pointed out.

The other option would be to create rosters based on a revised ratings system with the intention of focusing on realistic, on-field gameplay. The problem here is that means lowering player ratings substantially, in which case those questionable sim numbers you listed are going to be even less accurate.

It's a fundamental flaw with the game and it's been this way since 2K5. It has to do with the way the hitting ratings specifically interact with the game engine. And bottomline, one has the choice between rating/editing for accurate sim results or for realistic, on-the-field gameplay, but not really both.

With the default ratings system 2K uses, an inordinate number of hit balls end up in the air as fly balls, looping fly balls and line drives. That's because the power ratings for too many players are simply too high. But it's also because the contact rating acts as a loft modifier -- the higher the contact rating, the more loft the ball gets.

A typical power hitter in default 2K will generally have ratings along the lines of 90 power and 80 contact. Again, without the contact rating jacked up like that, the player isn't going to end up with the requisite number of HRs at the end of the season. But the end result in-game is that pretty well every time that hitter makes contact, it's going to be a fly ball of some sort. Groundballs, dribblers, etc. are extremely few and far between in this game, and always have been.

Editing for realistic, on-field gameplay essentially means re-scaling ratings for the specific purpose of seeing a lot more groundballs and soft hits. So instead of 90 power/80 contact, a power hitter might be rated 85 power/45 contact. Contact hitters in the game are generally rated something like 70 power/90 contact. That would change to maybe 35 power/85 contact. Etc.

It's then possible to create different and interesting kinds of scrub players. In typical unimaginative fashion, most scrubs in the game are simply assigned lower ratings in both catagories, as if everyone who's not a name player in real life are all anemic contact and power guys. Well, one can instead create light-hitting speedsters who tend to pop up a lot (10 power/75 contact). Or heavy hitters -- backup catchers for example -- who hit most everything on the ground (85 power, 20 contact). It makes things much more interesting in that these hitters are now all quite different, each with their own particular weaknesses and upside as well. And on the defensive side of things, one has to now think about shifting the defense to accomodate these different hitting styles and abilities, something that's generally not required with the vanilla ratings.

That's all for now.

[ksig24]

Re: Attributes and Ratings

This is interesting stuff Pei! Incoming friend request. I would love to chat more via Meebo.

[wudl83]

Re: Attributes and Ratings

There seem to be more flaws. Like some with speed ratings, where too many players are rated far too slow in my opinion. But when you increase their speed rating you got to know their overall rating will go up also. That creates "unrealistic" lineups and so on.

I have edited a MLB offseason roster for the pc version and have gone over to make it more realistic for played games. I do not see any sense in having power hitters, that don't hit vor average in real life, with high contact ratings in the game. Same goes the other way round.

Another big problem are the pitcher's ratings. Pitchers with high control in real life do tend to give up big numbers and many runs in played games because most lack movement in their balls. But when you up their movement it is nearly impossible in played games that you hit their balls.

Well, one must deal with it and I don't think that there are exact formulas for the ingame ratings since you can see that some players aren't rated based on their last years performance but on their average performance over 3-4 years.

I have gone over to something like that and used that as rough guidelines:
hitters:
contact rating = hitting avg
doubles rating = based on doubles/at bat
triples rating = based on triples/at bat
homerun rating = home runs/at bat
fielding = based on the uzr rating you can find at fangraphs
speed etc. = based on the speed rating you can find at fangraphs
eye rating = based on how much Ks a player gets and on Ks/BBs
K tendency = Ks/9 innnings
BB tendency = BBs/9 innings
stealing tendency = sb/at bats

pitchers:
movement = based on Ks/9 and BBs/9 and avg. a pitcher allows
control = based on Ks/9 and BBs/9
composure = LOB%
stamina = innings/game
contact = avg. allowed
doubles = always around the contact rating
triples = always 25
homeruns = hr/9 allowed
walk tendency = bbs/9
K tendendy = K/9

And so on.

[wudl83]

Re: Attributes and Ratings

What I want to add:
In my roster there are many many players with average ratings, which represents the MLB in my opinion much more than the original rosters. Also the most AAA players are rated between 60-70.

[peigone]

Re: Attributes and Ratings

Well, one must deal with it and I don't think that there are exact formulas for the ingame ratings since you can see that some players aren't rated based on their last years performance but on their average performance over 3-4 years.

In my roster there are many many players with average ratings, which represents the MLB in my opinion much more than the original rosters.

Good points in your post. I think you and ksig and anyone else who wants to create rosters the community will be interested in using, have to deal with the flaws and game engine imbalances by finding a happy medium so to speak. Sounds like you're doing that. I edit my roster for me, and I don't care about sim stats as I don't play career mode. But most people obviously want their end-of-season stats -- HR totals, etc. -- to at least look somewhat MLB-like.

ksig, it's always great to come across fellow gamers who understand what I'm even talking about, heh. The response I usually get is something along the lines of "Ugh, that's too much work, no way," or "We shouldn't have to fix the game like that, it's the devs' responsibility, huff puff!"

[mkharsh33]

Re: Attributes and Ratings

I just try to see how the sim engine stats play out and devise a "system" from that. It's a ton of work, and I'm not certain I'm going to do rosters this year (or I'll just do them for myself). Been doing them for 3 years and it's mostly complaints... I just don't have time to follow MLB players as closely as I'd like, so I'm doing NBA rosters because it's only a max of 15 players, and I understand basketball a lot better than baseball, and that there are less ratings for NBA players, which are often more obvious to define. Good conversation here, though...

[wudl83]

Re: Attributes and Ratings

Yah, I understand you Mkharsh.
It is a ton of work to make the rosters accurate because of the big amount of players, espacially when you want to edit the AAA, AA and A affiliates also.

But good point, I will sim a season tomorrow with my rosters. That shouldn't last too long since I am currently very busy.

One more problem I see is that no one knows how exactly the overall ratings are calculated. Overall ratings affect the lineups and the lineups affect simmed games. There are ratings which are more important than others, but I do not exactly know which ones.

And another problem: the talent stars/ratings which are far too high in the official rosters in my opinion.

It would be much easier, when 2k (and other sport series also) would reveale their "codes" so one could take them and do work based on them. I don't think that would be a problem for the series or developers since other games use other mechanics and 95% of the gamers don't care either. But the few editors woudl be very pleased...

[peigone]

Re: Attributes and Ratings

One more problem I see is that no one knows how exactly the overall ratings are calculated. Overall ratings affect the lineups and the lineups affect simmed games. There are ratings which are more important than others, but I do not exactly know which ones.

That's pretty easy to figure out. I do it all the time with games like FIFA and PES -- systematically test how each skill rating affects the player's overall, depending on their particular position. MLB 2K is much, much easier to test in that the contact and power ratings for all position players for example comprise almost all of a player's overall.

[peigone]

Re: Attributes and Ratings

I want to add that I don't put much stock players' overall ratings when editing. IMO that's half the problem with these games. There can be players who have a specific value and skill, but are rated lower than most others in the game -- someone with 97 speed, or a player with high contact from one side of the plate only.

PES is one of the only games that really tries to create useful average/below average players (or used to in the good old days). Rosters were full of lower-rated players who were better than their teammates in one or two very specific skill areas. So it was up to you the manager to know each player's particular strengths and weaknesses, and adjust the lineup accordingly.

The way most games are programmed these days is the devs rate all the name players high in most skill areas, and rate all the scrubs low in all skill areas. Very lazy and unimaginative game programming, and not very realistic either.

[wudl83]

Re: Attributes and Ratings

That's pretty easy to figure out. I do it all the time with games like FIFA and PES -- systematically test how each skill rating affects the player's overall, depending on their particular position. MLB 2K is much, much easier to test in that the contact and power ratings for all position players for example comprise almost all of a player's overall.

Ok thx. Well, that sucks.

[wudl83]

Re: Attributes and Ratings

I want to add that I don't put much stock players' overall ratings when editing. IMO that's half the problem with these games. There can be players who have a specific value and skill, but are rated lower than most others in the game -- someone with 97 speed, or a player with high contact from one side of the plate only.

PES is one of the only games that really tries to create useful average/below average players (or used to in the good old days). Rosters were full of lower-rated players who were better than their teammates in one or two very specific skill areas. So it was up to you the manager to know each player's particular strengths and weaknesses, and adjust the lineup accordingly.

The way most games are programmed these days is the devs rate all the name players high in most skill areas, and rate all the scrubs low in all skill areas. Very lazy and unimaginative game programming, and not very realistic either.

Exactly this.

A good example for a good player which is underrated ingame because he has no power and is quite slow is Freddy Sanchez. If I want to give him ratings that affect his real-life play it is IMPOSSIBLE to get him over a 78 overall rating, despite him having good contact (around 80) and good defense (way above 80 in most fielding ratings). Players like Kelly Johnson, Jason Kipnis, or even Orlando Hudson have a better overall rating ingame despite them being worse in real-life (in my opinion) simply because they are either faster and/or have more power in their bat...

[mkharsh33]

Re: Attributes and Ratings

Great comments...

What bothers me is when I try to get a guy to play to his strengths, but his OVERALL number isn't all that high. And then hearing people piss and moan about it, not realizing that I raise his overall, then I have to strengthen him in other areas that he's simply not good at...

[shogunofharlem3]

Re: Attributes and Ratings

Great comments...

What bothers me is when I try to get a guy to play to his strengths, but his OVERALL number isn't all that high. And then hearing people piss and moan about it, not realizing that I raise his overall, then I have to strengthen him in other areas that he's simply not good at...

I don't mean to chime in. You roster guys that put in all this effort are the berries, I'll start with that. Thank you!

I think overall rating is the single most overrated attribute of them all. Example, I just did a goalie edit for NHL 2k10. The highest rated goalie is a 76 I think. They play amazingly!

But when I first posted my edit results alot of first impressions were, HOW CAN TIM THOMAS BE A 76??? Well, he is a 76 rating on a scale of 76 due to attribute tweaks. It is all relative.

I have enjoyed reading this thread. Sort of got me interested in maybe playing around with some CAP in MLB 2k11. Great work boys! Keep it up!