MLB Reference Stats

[chriscf10]

I have always worked really hard on tuning baseball games to play as close to what you would see in the MLB. Let me say now, this is not a set of sliders. Rather, this thread will give you guys reference stats to compare your results from whatever sliders you are using. I have not shared a complete set of sliders since the days of High Heat Baseball when I would post my Tune file settings. I am just now beginning my slider testing after having used this past week or so getting used to the gameplay. I may post sliders after I'm happy with the results, but I know many of you are already hard at work doing the same.

Anyway, if you guys are using or tweaking sliders in efforts to get a realistic game, here are some stats that you will want to compare your results to. I used several websites to compile these stats. The first and most important aspect of any baseball game to me is the pitcher/hitter dynamic. If this is off, I have trouble enjoying the game.

Using the Corners/AI Corner
If you are just considering pitches that use the outer and inner 1/3 of the plate and are in the strikezone, then you are looking at 39% of all pitches. If you include pitches that are off the outside corner and off the inside corner, you are looking at 80% of all pitches.

% of pitches thrown for Strikes/AI Throw Strike Overall
Of all the pitches thrown last year, 62.7% were counted as strikes. However, this includes pitches that were swung at that would have actually been called balls. If you factor that in, the percentage drops to 51.4%.

% of pitches thrown for Strikes when pitcher is ahead in count/AI Throw Strike Ahead
55.8% if you include swings at pitches that would have been called balls and 41.6% if you go strictly by the strikezone.

% of pitches thrown for Strikes when pitcher is behind in count/AI Throw Strike Behind
69.4% includes swings at pitches out of the strikezone while only 59.2% were actually in the strikezone.

So for pitching, whatever sliders you choose to use, here is what you should compare your results to:
AI Corner: 39-80% of all pitches
AI Throw Strike Overall: 51.4-62.7% of pitches
AI Throw Strike Ahead: 41.6-55.8% of pitches
AI Throw Strike Behind: 59.2-69.4% of pitches

Now on to AI hitting.
AI Take Strike Overall
Hitters in the MLB took 33.7% of pitches that were in the strikezone overall

AI Take Strike Ahead
Hitters took 34.7% of pitches that were in the strikezone when they were ahead in the count.

AI Take Strike Behind
Hitters only took 16.5% of pitches that were in the strikezone when they were behind in the count.

AI Take Ball Overall
Hitters took 75.6% of all pitches outside the strikezone overall.

AI Take Ball Ahead
Hitters took 75.8% of pitches outside the strikezone when they were ahead in the count.

AI Take Ball Behind
Hitters took 70.4% of pitches outside the strikezone when they were behind in the count.

AI Contact
Hitters made contact with the ball on 81.5% of all their swings.

AI Foul Ball Frequency
Hitters fouled off 46.7% of pitches they made contact with.

AI Power
33.6% of all hits went for extra bases, while Total Bases/Total Hits was 57.7%.


As for some other stats that you guys may find useful:
Stolen Bases: Runners were successful on 74% of their stolen base attempts last year. 0.8SB were attempted per game.
Errors: 0.6/G per team, so 1.2/G overall

Another interesting one that seems to get tweaked is AI Runner Aggression. Here's what I do to track that. If a runner advances 1 base or more than the hitter, that counts as an extra base. I.e. a runner goes from 1st to 3rd or 2nd to Home on a single or goes from 1st to Home on a double. You can also factor in Sacrifice Flyes if you like. Considering all of that, runners took an extra base 57.3% of the time they had the opportunity to do so.


So there you have it. If you want a game that plays like the MLB, you should try to adjust your sliders to give you results like this. I'm going to begin tweaking my sliders now and testing them. However, the process usually takes me several weeks to get a set that I'm happy with. To do my testing, I start with exhibition games and only use #3 or #4 starting pitchers for each team. These are generally the "average" pitchers and should give you the closest results to the averages I listed above. If you use and play against Aces or #5SP's, your numbers will be skewed. I also use and play against as many teams as possible. This will give me my framework of results. I then take that into a Test Franchise to make sure the results hold. If they do, I start my real franchise.

I hope you slider tweakers can use this info. Have fun with it!

[aj719887]

Re: MLB Reference Stats

some interesting stuff here... i would love to get my sliders as close to these as i can!!

[Amazin]

Re: MLB Reference Stats

You would think 2K would actually base a sim baseball game on some actual research like this but I highly doubt it.

[DIESEL]

Re: MLB Reference Stats

Very interesting stuff. Thank you.

[arichzona]

Re: MLB Reference Stats

thx for the research man........... saved me some time. I am still tweaking myself to get it to play as close as possible - trouble is you move one slider, one notch, and the whole thing can go to Hell.........or Heaven -hopefully...........sometime soon.

Of course, I want realism but.... I do not want my games taking 2 hours either.

[chriscf10]

Re: MLB Reference Stats

I am very close to having my sliders set to give a realistic game. My test franchise has been with the Mariners and I have been within 3-5% of their actual stats from last year with the exception of extra base hits. I'm getting too many, but I have made an adjustment and will be testing it this next week. If all goes well, I will have a set to post late this week or early next week. This takes a while since I test these by playing many games making sure I use all players and pitchers and making sure I face a variety of teams and opposing pitchers. The only stat that I can't seem to get a fix on is the CPU stealing as they just don't attempt enough.

[baa7]

Re: MLB Reference Stats

% of pitches thrown for Strikes/AI Throw Strike Overall... 51.4%.

That's not going to happen in this game, not even with all three throw strikes sliders at 0. They're simply not effective enough this year, unlike in other versions of 2K baseball.

The only way to get close to this number is by dropping pitchers' Strikeouts and Walks Allowed ratings way down. In the tests I've done, I'd guess those two ratings have to be lowered by a minimum of 60% for the majority of pitchers. And when they are lowered, pitchers do start throwing an equal number of balls and strikes, or close to it anyway.

[chriscf10]

Re: MLB Reference Stats

That's not going to happen in this game, not even with all three throw strikes sliders at 0. They're simply not effective enough this year, unlike in other versions of 2K baseball.

The only way to get close to this number is by dropping pitchers' Strikeouts and Walks Allowed ratings way down. In the tests I've done, I'd guess those two ratings have to be lowered by a minimum of 60% for the majority of pitchers. And when they are lowered, pitchers do start throwing an equal number of balls and strikes, or close to it anyway.

The upper range of the average pitcher is 62.7%. Without altering any player ratings, I am seeing the AI throwing right at 62% of their pitches for strikes. The actual number is a bit lower as I have swung at pitches that would have been called balls (you can track this if you swing and miss by waiting for the final pitch location to show up and just a second later, it shows what the ump would have called). The more games you play in franchise mode, the more variable the umps are. I've had games where the AI threw over 70% strikes and some where they just get over 50%. This is largely due to some umps having a large strike zone and some have a small one.

The most important thing to remember for any person's slider setting is their style of play. Any one person's sliders won't give identical results for another person because their style of play is different. When I post my sliders, I will give a detailed description of exactly how I play this game down to individual batter counts and various game situations. If you don't know how a person plays the game, you just can't get an idea of if the sliders will work for you without spending a ton of time with them. By posting how I play, hopefully people will be able to simply look at my style to see if it matches how they play. If so, they should feel good about trying them. If not, they should not waste their time. I'm not going to be so bold to say my sliders are perfect or even good for everyone. That would be tremendously foolish. What I will say is that if you play this game similar to the way I play it, I think you will be happy with the results you get with the sliders if you are going after actual MLB results.

[baa7]

Re: MLB Reference Stats

The upper range of the average pitcher is 62.7%. Without altering any player ratings, I am seeing the AI throwing right at 62% of their pitches for strikes.

Really? That must have to do with the umps as you said, although one would think it would balance out (umps with wide strikezones versus tight). I'll keep track a little closer, but as I said I don't believe I'm seeing anything like that in the exhibition games I'm playing.

Great thread, BTW

[baa7]

Re: MLB Reference Stats

AI Contact
Hitters made contact with the ball on 81.5% of all their swings.

That's an amazing stat. If that had been a trivia question, I would have answered less than 50%. My foul ball slider needs to be bumped up, because CPU hitters aren't touching the ball anywhere near that %.

[baa7]

Re: MLB Reference Stats

BTW, regarding my comments about the throw strikes sliders: In previous versions of 2K baseball, you could lower those sliders to 0 and the CPU would walk every single batter in your lineup. And I don't mean most of your batters: I mean the CPU pitcher would not be able to find the strikezone, period.

It's great if the game can come close to matching those MLB stats with the throw strikes sliders at 0. But my point is that in previous years, it wouldn't even be an issue. In previous years, you would have to bump those sliders up into the 40 range, if you wanted to match those stats. And that's the way these sliders should be programmed.

One problem I have with the game this year is that a number of the sliders don't appear to be properly programmed, while some appear to be completely broken (steal sliders). For example, if you want to see how poorly the individual pitch speed sliders are programmed, check out my thread in the rosters section:

http://www.operationsports.com/forum...ch-speeds.html

[chriscf10]

Re: MLB Reference Stats

BTW, regarding my comments about the throw strikes sliders: In previous versions of 2K baseball, you could lower those sliders to 0 and the CPU would walk every single batter in your lineup. And I don't mean most of your batters: I mean the CPU pitcher would not be able to find the strikezone, period.

It's great if the game can come close to matching those MLB stats with the throw strikes sliders at 0. But my point is that in previous years, it wouldn't even be an issue. In previous years, you would have to bump those sliders up into the 40 range, if you wanted to match those stats. And that's the way these sliders should be programmed.

One problem I have with the game this year is that a number of the sliders don't appear to be properly programmed, while some appear to be completely broken (steal sliders). For example, if you want to see how poorly the individual pitch speed sliders are programmed, check out my thread in the rosters section:

http://www.operationsports.com/forum...ch-speeds.html

I agree with you regarding the slider sensitivity. Some seem to be very sensitive while others tend to need huge changes to see a noticeable change. Like you said, the AI just does not steal no matter while if I were to put the sliders I'm using in this year's game into 2K7, I would have hated 2K7. I actually started with the sliders I had in 2K7 and quickly determined that was not the way to go. So, I went from the ground up.

Regarding the strike %, I too had the same initial reaction as almost everyone who played an exhibition game, the AI just does not throw balls. But once I started my test franchise, I found that the more games you play, the more things comes back to some level of normalcy. I have really enjoyed the variability of the umps in this game. I have come across some who would give a good inch or 2 on the corners or high and low. The ball icon wouldn't even be touching the strike zone. I really used that to my advantage to get tough hitters out. On the other hand, I faced the Yankees against an ump that wouldn't call a strike unless more than 75% of the ball icon was in the strikezone. Needless to say, I got pounded, but I also pounded their pitcher. I have found that I'll go on a run of several games where I think the stats are going to get out of whack, but they always seem to normalize the more I play. I had a few times where I thought I had the AI hitting too powerful, but after taking a couple of beatings, I bounced back with a few well-pitched games. This game seems to ebb and flow very nicely. I have played better baseball games, but this one will do since it is my only option on the 360.

[chriscf10]

Re: MLB Reference Stats

That's an amazing stat. If that had been a trivia question, I would have answered less than 50%. My foul ball slider needs to be bumped up, because CPU hitters aren't touching the ball anywhere near that %.

Don't get too hung up on this stat. You just want to make sure your K/9IP, BB/9IP, and Total Hits are believable.

If you really want some trivia:
Tony Gwynn 1994
100% contact rate on 2-0, 3-0, 3-1, 3-2 counts
Overall, he only swung and missed 48 times that year (strike season). His average was just around 70 swings and misses for an entire season. That's insane!

Ichiro swings and misses 2x that on average, but still makes contact 88% of the time.

Adam Dunn swung and missed at 322 pitches last year for a 71% contact rate.

Ryan Howard was even worse at 418 whiffed pitches for a contact rate of only 64%.


As for pitchers, batters only made contact with Jake Peavy pitches 76% of the time, Johan Santana 72%, Erik Bedard 74%. However, Carlos Silva had 87% of his pitches that were swung at hit.