Please take a good hard look at POWER vs CONTACT balance

[JeffersonJim]

Contact = Power & AvoidK & BABIP (Secret)

Disclaimer: this is pretty oversimplified as I know it. The BABIP factor is a hidden stat to players. Overall Contact is like squeezing a balloon filled with Power, AvoidK, and BABIP. The more power usually equals less BABIP and/or AvoidK. If the player already has very high Power and the contact is low, they will probably not have a very high BABIP as a result, and hit for a low average. Also understand these factors are NOT in equal portions and not necessarily added together. Most want to know the magic top secret formula.

If the player has high Contact (say 90+), but low Power (below 50), this player will usually generate higher BABIP as a result. If they also have higher AvoidK, their total OBP will be high. Of course this does not mean they will generate high WPA, wRC+, wRAA or even WAR.

Last, my experience over the past few years is many Diamond/Perfect teams ballparks are decreased in HRs much more dramatically than AVG. So a team more built for BABIP can better take advantage than a team more built for HRs, more often

[pappyzan]

Originally Posted by Orcin

Yes, although it is probably not that simple. But the total number of home runs per league per year is fixed.

Well, it's not fixed as in the league will hit only X number of HRs each season, or pitchers will throw only X number of Ks each season. League totals represent a range where totals fall between X and Y.

X doesn't mark the spot, but the spot is somewhere between X and Y. X and Y are based on a particular season for the game engine, last year this was 2010 from what I recall.

League totals aren't a unique concept within only PT, it's how it works in OOTP base game as well.

[chazzycat]

I wonder also if the eye ratings are not part of the problem.

Barry Bonds at his peak was walking ~30% of the time, while the average walk rate in MLB was 9%.

Mike Trout's career walk rate is ~15%, while the average walk rate in MLB has not changed meaningfully since Bonds era.

So how does PEAK Bonds & LIVE Trout have identical eye ratings? (102 for Trout, 101 for Bonds)

Ruth is a similar story. Career walk rate of ~20% when the MLB average was around 8%. So again he should be far better than Trout compared to his peers, and yet he only has 3 points over Trout.

[EvvCat]

This is a great thread. I hadn't really spent time thinking about how the engine being built around recreating specific season totals – which works *great* for replaying standard seasons or leagues made of a year's worth of players of all skill levels rearranged – affects results in Perfect Team, where the same player can be on 30 teams and every player is strong (relative to average MLB players of whatever year/era).

Thinking it through, this *really* dampens my enthusiasm for PT. I guess the engine might work OK in the PT world if teams are balanced, and the concept of promotion and relegation works toward that over time, but this capping of totals to an established range exposes some inherent flaws in applying a season-recreation model to an open world where every team can have any player.

When I picture a league where every team has only pitchers rated in the 40s and lineups of hitters all rated 95-100 and realize that league would still finish with about the same number of home runs, strikeouts and everything in between as a balanced league... Well, I don't know how to reconcile that and still feel like the roster and strategy decisions I make will have the impact I'm supposed to believe they'll have.

Am I missing something that will keep PT feeling honest?

[chazzycat]

I honestly don't have any problem whatsoever with the fact that league normalization exists. I struggle to imagine any other system for keeping things even close to believable. To me that is not the problem...it just needs some fine tuning.

[Spieler]

Something changed this year, actually two technically. Stuff and Power had their caps taken off (i.e. bumped from max 128 to 256). Surprised no one has mentioned that.

[neugey]

I do agree, the ratings could use some tuning so things balance out a bit more and contact is offset better by the others. And actually a bigger concern of mine is that doubles/triples park factor aren't held to the same threshold the AVG/HR is at .9-1.1. I wonder if that is skewing things as well.

But at the end of the day, play Bonds and Ruth if you like; winning isn't necessarily the be-all end-all seeing as this is not an e-sport and no one is playing for real world prize money.

Perhaps if we could get a bit longer of a PT closed beta period for PT22, this is something we could help the dev's test out and find the balance.

[chazzycat]

Originally Posted by Spieler

Something changed this year, actually two technically. Stuff and Power had their caps taken off (i.e. bumped from max 128 to 256). Surprised no one has mentioned that.

Yeah, this is a good point. I was hopeful when they announced this, that it would change things from last year. But it seems we have enough data at this point to say, it didn't.

I would venture to guess that they basically cancel each other out, since HR sluggers have bad AvoidK ratings. Oversimplified but I think that's the gist of it

[Spieler]

Now this is far from scientific (on purpose), but it is common knowledge that CON is a combination of HR, K and BABIP. If 2 of those 3 have factors that (oversimplified) can go up to 256, leaving the third one alone at 128, whaddya think is going to happen?

[Lemandria]

Try to remember that 2019 (the season on which these cards are largely based) was a juiced ball year, which would tend to devalue homers generally. Trashcans.

And (as we've seen in XX) really boost the value of MOV and devalue STU relatively.

What doesn't feel "right" may not be the sort of year baseball had, in any given era.

[chazzycat]

Yep. I think if they set the run environment in PT to 2019, the pendulum would swing entirely in the other direction, and Bonds/Ruth would be god-tier cards. Not that we want that, either.

[One Great Matrix]

Originally Posted by dkgo

Its not the lack of homers since every league will have the same number hit in a season and Ruth will always get his share, but the way the ratings work when two players have similar contact then one with very high power will hit in the toilet for average

I think, anyway, that this is because of their low avoid K's...and the power K pitchers at the high levels, not because of their high slug. Make sense? They strike out more often, and they are facing Randy Johnson's and Dwight Gooden's, Nolan Ryan's, the other guys that aren't ummm soooo commonly known as guys who could fool a slugger out his shoes.

[One Great Matrix]

Originally Posted by chazzycat

I'm not trying to wade into a philosophical debate about the relative merits of contact vs. power in real life baseball. Certainly, players of both varieties have left their mark on the game throughout history. So I strongly believe that both types of players should be viable options in PT.

It's just a matter of balancing the game properly. If one side is too obviously strong there's no tradeoff to consider, no interesting decision to make. Just take the contact & range every time. It's boring.

What do you mean? What if they mostly do a good job of making the players digital representation based on the data, & maybe if they faced aces all season long, then things would go somewhat more similar to the way they do...I mean are you a baseball fan? Not saying it in any degrading way, but I notice that out of any sport baseball leaves me in disbelief sometimes with expected vs. reality. And this, although a fantasy...perhaps is doing a good job of demonstrating this. Not "off".

[One Great Matrix]

Originally Posted by chazzycat

This is a common misconception. With the way league normalization works, you should be more concerned about the other power hitters in the league, not the pitchers.

Don't get me wrong, MOV does prevent homers. So if for example you have a bunch of teams in your division with huge MOV ratings, it will hurt you.

But generally speaking, in your fight for the slice of the HR pie, it's the other power hitters in the league you're competing with. Not the pitchers.

You said nothing.
Thanks.

[allenciox]

So, chazzycat is right on that the POW rating is part of CON. Also AvoidK is part of CON. The third rating that is a part of CON (the most important part) is BABIP.

Everything that happens in the game is based on "opposition rolls", i.e. a roll that takes into account a single rating for an "offensive" player versus a single rating for a "defensive" player. The probability required for a roll to work is based on this opposition roll adjusted for league normalization and park normalization.

For a walk, the opposition roll is Pitcher CON versus Batter EYE.
For a strikeout, the opposition roll is Pitcher STU versus Batter AvoidK.
For a home run, the opposition roll is Pitcher MOV versus Batter POW.

Ok, everybody probably knew that already. But for a hit vs. an out, the opposition rating is

Batter BABIP vs. Fielder Range (to determine whether fielder gets to a ball), or, in some cases, Fielder Arm.

So the pitcher has very little role in determining whether a hit ball is an out, a single, or a double/triple (double vs. triple is purely dependent on batter speed .versus park considerations)

In OOTP 21, you have great fielders at all levels of ability. The difference in fielding between the best bronze players and the best perfect players is virtually zero. (consider 68 Ozzie Smith for example). However, the ratings for POW and CON are generally much higher as the OVRs go up.

So the result is that, as you get higher OVR players across a league, the AvoidK component of CON becomes "more" important, and the POW component becomes "less" important, while the BABIP component stays roughly the same in importance.

The result is that somebody like SE 92 Ichiro is going to get very similar batting averages as they go up the ranks, since the AvoidK becomes more important at higher levels, while his low power is going to have minimal impact on BA (maybe he hits only 4 HR a year instead of 10 HR --- six less hits; while a high POW guy might get 20 HR vs. 50 HR, or 30 less hits at the highest levels).

On the other hand, somebody who has the same CON rating as Ichiro but has very high POW and relatively low AvoidK, is going to have BA decrease significantly as they ascend levels.

Let's take a real-life example from one of my teams.

SE Mike Piazza (96) has 95 CON, 108 POW, 67 AvoidK
SE Ichiro Suzuki (92) has 96 CON, 58 POW, 86 AvoidK

Piazza, in 5 years of Diamond League play, has a lifetime .237 average
Ichiro, over the same period, has a lifetime .309 average

The difference is incredible, even though their CON is almost exactly the same.

Now Piazza might get some decrement since my park factors favor lefties. Let's take a different example. On the same team, I have lefty Ted Williams (100) at DH. CON: 100, POW: 119, AvoidK: 70.

His career Diamond stats: .274 BA, even though his AvoidK is still pretty good, and his CON is higher than Suzuki's.

[chazzycat]

Originally Posted by One Great Matrix

What do you mean? What if they mostly do a good job of making the players digital representation based on the data, & maybe if they faced aces all season long, then things would go somewhat more similar to the way they do...I mean are you a baseball fan? Not saying it in any degrading way, but I notice that out of any sport baseball leaves me in disbelief sometimes with expected vs. reality. And this, although a fantasy...perhaps is doing a good job of demonstrating this. Not "off".

If you insist on bringing it back to real baseball and not talking about game design, here's some interesting info: https://sports.yahoo.com/mlb/players...r&type=Batting

Here are some of Bonds's OBPs versus various elite pitchers of his era:

- Maddux: .370
- R. Johnson: .452
- Glavine: .442
- Schilling: .410
- Gooden: .440
- Smoltz: .467
- K. Brown: .483
- Martinez: .448
- Saberhagen: .429

But do go on about how his skill set doesn't translate against elite pitching.

[OMGPuppies]

Originally Posted by chazzycat

If you insist on bringing it back to real baseball and not talking about game design, here's some interesting info: https://sports.yahoo.com/mlb/players...r&type=Batting

Here are some of Bonds's OBPs versus various elite pitchers of his era:

- Maddux: .370
- R. Johnson: .452
- Glavine: .442
- Schilling: .410
- Gooden: .440
- Smoltz: .467
- K. Brown: .483
- Martinez: .448
- Saberhagen: .429

But do go on about how his skill set doesn't translate against elite pitching.

I don’t see your point.

[pappyzan]

Originally Posted by OMGPuppies

I don’t see your point.

Yeh thanks for nothing Chazzycat

[Catchthedamnball]

Originally Posted by pappyzan

Yeh thanks for nothing Chazzycat

Go back to spamming your bo3s. Jeez.

[Syd Thrift]

Originally Posted by Spieler

Now this is far from scientific (on purpose), but it is common knowledge that CON is a combination of HR, K and BABIP. If 2 of those 3 have factors that (oversimplified) can go up to 256, leaving the third one alone at 128, whaddya think is going to happen?

Even though the game presents you with a Contact rating, that rating is not used directly by the game. Basically, so far as I know, the game flow looks like:

- determine if a player struck out
- if not, determine if they walked
- if not, determine if they hit a HR
- if not, determine if their ball in play was a hit or not

No matter what the ceiling of any of those are, you’re not going to have a stack overflow when you “combine” them because you’ll never combine those ratings. There’s some sort of “stuff vs avoid Ks” kind of mojo that occurs but IME you don’t see stack overflow issues there either, and there are a variety of ways to avoid that, and I don’t know exactly how the game makes those “rolls” I noted above in the first place, so I’m not even going to speculate on how.

Just, as the big takeaway, Contact is not directly used by the game, only its component parts.