MLB Player overall ratings - perception versus reality

[CBLCardinals]

MLB Player overall ratings - perception versus reality


Hey gang,

Baseball America, one of the best baseball resources out there, had a great article last week for the beginning of the MLB season that listed each teams projected lineup, starting rotation, and top three bullpen arms. This is fairly standard stuff starting a new season….however What I’ve not seen before is they also placed current overall grades on all of the players.

Below are the results rolled up by grade.

Some thoughts…

- I LOVE to see OOTP’s direction get closer and closer to this model to reflect reality
- Most MLB players are somewhere in that 45-55 range (nearly 77%)
- Top RP max out at ~60 per Baseball America. In this study that would be Edwin Diaz and Josh Hader. Historically, someone like Mariano Rivera would be a 65+ perhaps. OOTP is high on RP grades versus reality IMO.

For those that prefer a bajillion 70 and 80 grade prospects in drafts that is just not simply based in reality looking at current MLB grades. As is, MLB currently has 28 players in the entire league that rate 70+.

Lukas also brought up a great point in another thread that not all prospects develop in real life or in OOTP.

Also…. at some point we have to realize a ~50 level player is around average and a valued player. You need lots of average players that can supplement a roster and avoid massive holes in a lineup.

Using these understandings above has been the basis for the creation of my 2024 MLB roster set. I have ~33 current rating 70+ either SP or batters (not counting IL) and Real MLB per Baseball America has 28.

My current calibrated 2024 MLB draftees have 7 players with potential of 55+, and 29 with 50 potential.

Just wanted to take some time and share my insights and thoughts around all things ratings related.


If anyone would like to try the roster set out, link is below. Look for the last post from me for latest version.

https://forums.ootpdevelopments.com/...d.php?t=354076

[Jwalk100]

Thanks for this!

[snepp]

Out of curiosity, the current ratings distribution in my 28-team fictional on opening day. I've run enough seasons to have all players generated on most recent update.

[Paul Reuschel's Mustache]

Originally Posted by CBLCardinals

- I LOVE to see OOTP’s direction get closer and closer to this model to reflect reality
- Most MLB players are somewhere in that 45-55 range (nearly 77%)
- Top RP max out at ~60 per Baseball America. In this study that would be Edwin Diaz and Josh Hader. Historically, someone like Mariano Rivera would be a 65+ perhaps. OOTP is high on RP grades versus reality IMO.

Thank you for this!

I'm still getting used to the new ratings and it's posts like this that help me to understand why changes were made.

[Daniel_09]

I think we got used to previous versions of having in our roster about 15 or 20 players well above average player, and not seeing the reality of MLB in which a player considered 55/80 is a contender year after year to be in the all-star team, and that the regular players are between 40-50, having some relievers, emergency starters and bench players in the 30-35.

Not to mention that baseball America's numbers only include 3 relievers per team and only the 9 starters of each team's lineup, if we include the entire roster, it is possible to have close to 50% of the players in the 35-50 range.

[Daniel_09]

Originally Posted by snepp

Out of curiosity, the current ratings distribution in my 28-team fictional on opening day. I've run enough seasons to have all players generated on most recent update.

excellent results we are talking about 60%-70% of the total number of players between the range of 40-50 out of 80 players in MLB.

[CBLCardinals]

Originally Posted by Daniel_09

I think we got used to previous versions of having in our roster about 15 or 20 players well above average player, and not seeing the reality of MLB in which a player considered 55/80 is a contender year after year to be in the all-star team, and that the regular players are between 40-50, having some relievers, emergency starters and bench players in the 30-35.

Not to mention that baseball America's numbers only include 3 relievers per team and only the 9 starters of each team's lineup, if we include the entire roster, it is possible to have close to 50% of the players in the 35-50 range.

Good post!! Agreed with past versions!

[Larryk007]

Great post!!! I think my sim is much higher, so I need to take a look and see what is the problem!

[Syd Thrift]

Yeah with the 20-80 ratings in particular the devs made a concerted effort to base them on standard deviations the way Fangraphs and also Baseball America do. That means that with a major league average grade of 50, roughly 2/3rds of the guys in the major leagues should be between 40 and 60. I think OOTP ratings are a little more stretched out than what you'd expect in a normal distribution so there are probably more than like 1 or 2 players per 30 team league who are 80s in the current system but mostly it tries to achieve that.

Also of course there's the issue that distribution in baseball is more like the very, very right side of a normal-distribution bell curve; the actual average person in terms of baseball talent would probably have a career closer to Shaq Thompson's, if not worse:

https://www.baseball-reference.com/r...d=green-002sha

[jcard]

1) Thanks for sharing

2) One issue with comparing MLB distribution to prospect potential is that people have different preferences for what a potential rating should be trying to capture (median, highest, etc.). There is also the related question of whether their should be a confidence or variance metric that gives additional context to the potential rating.

3) There is also the “game” factor here. Really, in OOTP your team-building is more heavily on a simple talent evaluation than is a real-life general manager’s (and baseball already lacks system and player suitability / compatibility, and so forth that you find in other sports and make games like Football Manager real rabbit-holes for those inclined). For example, high or low ball hitters, pitching staff arm slot and pitch repertoire diversity, etc., are not considerations. So at some point having 70+% of a top 100 prospect list share the same rating (as in the case of MLB’s list) can make deciding among them in OOTP feel arbitrary. It is just a trade-off between faithful representation of something in reality and playability of that mechanic within the limitations of a game that is inevitably an incomplete model of that reality.

4) Enjoyed the material and comments. Thanks again.

[Tesla51P]

Originally Posted by jcard

1) Thanks for sharing
2) One issue with comparing MLB distribution to prospect potential is that people have different preferences for what a potential rating should be trying to capture (median, highest, etc.). There is also the related question of whether their should be a confidence or variance metric that gives additional context to the potential rating.

This is valid feedback. There are many like me that really enjoy the more realistic side of the game, but some people want the game to be more of a game. I think that having pre-loaded OOTP settings groups that allowed the players to set whether they want Future Value (median) versus max

potential to be their method would help groups of players happy.

[Tesla51P]

Thanks for sharing CBL. It is interesting to see how BA rates the league for comparison. Also, I can't say thank you enough for the amazing roster set. The default OOTP roster is definitely getting more toward a true 20/80 scouting grade approach, but your set does an even better job.

[OutS|der]

Ratings have always been inflated in sports games. No one wants to draft 65 overall players and think you're doing great and only seeing an 80 once a decade or so.

We see a 65 Reliever as being worse then an 80 Position player just due to the lower number, not realizing that even Mariano Rivera the best ever would never be graded higher then 65 as he'll never be as valuable as that position player.

It's not easy adjusting to reality, we're used to trying to build teams of all 99 overall players in Madden and other sports games. In most of these games the higher overall is all that matters, where in RL most players are stuck together overall talent wise but it's the individual talents that separate them.

[highandoutside]

Originally Posted by jcard

1)

3) There is also the “game” factor here. Really, in OOTP your team-building is more heavily on a simple talent evaluation than is a real-life general manager’s (and baseball already lacks system and player suitability / compatibility, and so forth that you find in other sports and make games like Football Manager real rabbit-holes for those inclined). For example, high or low ball hitters, pitching staff arm slot and pitch repertoire diversity, etc., are not considerations. So at some point having 70+% of a top 100 prospect list share the same rating (as in the case of MLB’s list) can make deciding among them in OOTP feel arbitrary. It is just a trade-off between faithful representation of something in reality and playability of that mechanic within the limitations of a game that is inevitably an incomplete model of that reality.

Interesting point, thanks.

I play mostly fictional and with most ratings off (I use stats), but when I'm faced with difficult draft or trade decisions between players who look very similar I've often look at the player personality traits and the compatibility between a player's personality and my staff (manager, pitching coach and hitting coach) to help with my decision. Also, I'll take into account pitcher tendencies - high K or high FB or GB pitchers - and hitter type (pull, neutral) and align them with my team and ballpark strengths.

(I have no idea how much personality traits and player staff compatibility really matter for development, but it seems to me a guy high in work ethic should stand a better chance to develop than with a guy with low work ethic.)

So, I think there are some nuances you can use to distinguish between closely rated guys, but your larger point is right: it is difficult to account for all the distinctions that matter in real life.

[glenstein]

Originally Posted by CBLCardinals

For those that prefer a bajillion 70 and 80 grade prospects in drafts that is just not simply based in reality looking at current MLB grades. As is, MLB currently has 28 players in the entire league that rate 70+.

Completely agree, but with a caveat. I do think OOTP has in the past had a LOT in the way of prospects with crazy potential ratings. So far as I can tell, this year is somewhat different, and my version of OOTP25 is really stingy with the upcoming draft, mostly listing 3 star guys even at the very top. So maybe something has changed? I have only done a very cursory look so I could be wrong and heck, maybe I just have terrible scouting.

However, if the idea is that there's a top tier of super-projectible prospects, but they also have a very high fail rate, it can make sense to "overshoot" on high potential guys. Though I might just be stating the same thing as Jcard. In terms of what we "should" go for, it's a fascinating question with no obvious right answer imo. My very opinionated answer would be to go for highest, and the likelihood of realizing that outcome (perhaps a bust rate) could be reflected in some other way.

Anyway, overall I want to say thanks because projects like yours breathe life into the game. Thank you for creating this roster set!

[glenstein]

Originally Posted by jcard

3) There is also the “game” factor here. Really, in OOTP your team-building is more heavily on a simple talent evaluation than is a real-life general manager’s (and baseball already lacks system and player suitability / compatibility, and so forth that you find in other sports and make games like Football Manager real rabbit-holes for those inclined). For example, high or low ball hitters, pitching staff arm slot and pitch repertoire diversity, etc., are not considerations. So at some point having 70+% of a top 100 prospect list share the same rating (as in the case of MLB’s list) can make deciding among them in OOTP feel arbitrary. It is just a trade-off between faithful representation of something in reality and playability of that mechanic within the limitations of a game that is inevitably an incomplete model of that reality.

Wholeheartedly agree! Both with the fact that (1) deciding between two guys with equivalent ratings in OOTP can feel arbitrary and (2) I think you are so so right about how GMs evaluate skills.


The way I would say it (and this gets into MLB as well as prospects) general managers get into opinionated differences on how to value skill profiles. And that can add a huge amout of fun. What did the Rays see in Fernando Rodney when they grabbed him off the scrap heap and turned him into an ace reliever? How is it that Dan Dombrowski had such a great run of "scouting" MLB level talent to see that guys like J.D. Martinez were the real deal?

It's a frontier that, in my opinion, would hold a lot of promise in terms of contributing to the fun of a game centered on team building and valuing players.

[jcard]

Originally Posted by highandoutside

Interesting point, thanks.

I play mostly fictional and with most ratings off (I use stats), but when I'm faced with difficult draft or trade decisions between players who look very similar I've often move to player personality traits and the compatibility between a player's personality and my staff (manager, pitching coach and hitting coach) to help with my decision. Also, I'll take into account pitcher tendencies - high K or high FB or GB pitchers - and hitter type (pull, neutral) and align them with my team and ballpark strengths.

(I have no idea how much personality traits and player staff compatibility really matter for development, but it seems to me a guy high in work ethic should stand a better chance to develop than with a guy with low work ethic.)

So, I think there are some nuances you can use to distinguish between closely rated guys, but your larger point is right: it is difficult to account for all the distinctions that matter in real life.

Agreed that these are nuances that SHOULD be considerations in OOTP. I was trying to keep my word count down, but I will briee see fly explain why I omitted acknowledging them. Basically, I have never seen confirmation that these factors do in fact change the meaningful results of the game engine. For example, I myself do not know that if two RHH with identical Power ratings and platoon splits are teammates whose park has a short left porch and one is a spray hitter and the other is an extreme pull hitter that the latter will hit more HR’s. In other words, I do not know if the pull tendency actually leads to more left field HR checks, or is used instead as a post-determination (of the PA) distribution variable. Perhaps the batted ball tendencies are active additional determinants of results alongside the main ratings. On the other hand, could they be traits generated to correlate with certain main attribute profiles during the player creation process and really provide only flavor text to resolved events (somewhat analogous to the fact that in SOM a ground-ball pitcher has the same distribution of fielder ratings checks among the positions as does a fly ball pitcher—the results on his card do not make any additional use of his infielders’ defensive ratings than those on the fly ball pitcher’s card)? Admittedly, I also have seen no refutation of the ostensible purpose, either. I suppose that in general I lean more towards skepticism in the absence of proof than assumption in the absence of rejection. In any case, I do keep an open mind to additional information. Apologies for the verbosity.

[locuspc]

Originally Posted by glenstein

Completely agree, but with a caveat. I do think OOTP has in the past had a LOT in the way of prospects with crazy potential ratings. So far as I can tell, this year is somewhat different, and my version of OOTP25 is really stingy with the upcoming draft, mostly listing 3 star guys even at the very top. So maybe something has changed? I have only done a very cursory look so I could be wrong and heck, maybe I just have terrible scouting.

Are you still early in a real world save? The way the game uses its real world prospect list means the second and third drafts in a real world save are extremely low quality, because the cream of the crop gets skimmed in the first draft but it doesn't generate new prospects because there's still enough junk players to fill out the draft with. Once the badly-designed/buggy drafts are out of the way I am seeing plenty of high-potential talent in the draft.

[Pelican]

Does anyone have a feel for how the ratings on the classic 20-80 scale translate onto the 0-100 scale I use? My impression is that players are still clustered in the middle, pursuant to a standard deviation scheme. Players with ratings above 80 are rare. However, at the low end, young players, there are plenty of single digits, even if their potential is much higher. [This makes sense to me. Most of these guys will never even sniff the major leagues.]. Does that correspond to how players are distributed on the 20-80 scale? I assumed (always a mistake) that the 1-100 scale simply "stretched" things out, in order to cover a decimal approach. Was I wrong about that?

[PSUColonel]

Originally Posted by CBLCardinals

MLB Player overall ratings - perception versus reality


Hey gang,

Baseball America, one of the best baseball resources out there, had a great article last week for the beginning of the MLB season that listed each teams projected lineup, starting rotation, and top three bullpen arms. This is fairly standard stuff starting a new season….however What I’ve not seen before is they also placed current overall grades on all of the players.

Below are the results rolled up by grade.

Some thoughts…

- I LOVE to see OOTP’s direction get closer and closer to this model to reflect reality
- Most MLB players are somewhere in that 45-55 range (nearly 77%)
- Top RP max out at ~60 per Baseball America. In this study that would be Edwin Diaz and Josh Hader. Historically, someone like Mariano Rivera would be a 65+ perhaps. OOTP is high on RP grades versus reality IMO.

For those that prefer a bajillion 70 and 80 grade prospects in drafts that is just not simply based in reality looking at current MLB grades. As is, MLB currently has 28 players in the entire league that rate 70+.

Lukas also brought up a great point in another thread that not all prospects develop in real life or in OOTP.

Also…. at some point we have to realize a ~50 level player is around average and a valued player. You need lots of average players that can supplement a roster and avoid massive holes in a lineup.

Using these understandings above has been the basis for the creation of my 2024 MLB roster set. I have ~33 current rating 70+ either SP or batters (not counting IL) and Real MLB per Baseball America has 28.

My current calibrated 2024 MLB draftees have 7 players with potential of 55+, and 29 with 50 potential.

Just wanted to take some time and share my insights and thoughts around all things ratings related.


If anyone would like to try the roster set out, link is below. Look for the last post from me for latest version.

https://forums.ootpdevelopments.com/...d.php?t=354076

Absolutely agree with all of this!! Great stuff!