The SP/RP model in OOTP needs to change

[Righty Groove]

Hello everyone,

I believe the Starting Pitcher/Relief Pitcher dynamic in OOTP needs major improvement. Right now, it's basically about 2 things: Stamina and having more then 2 pitches. But it should be so much more.

First, most most MOST relievers are failed starting propects. They either didn't have the quality (either stuff or control), endurance or durability to start. Relieving is very different to starting, but it's safe to say it's considerbly easier too.

When "converting" a starter to reliever, not only should he get a boost in stuff, but also (and most importantly) in control and velocity.

A lot of pitchers get relegated to the bullpen for lack of control, for too high of a walk rate. Getting a boost in control as a reliever would adress that. Throwing harder is also a know "side effect" of relieving, with plenty of examples. Throwing high-90s as a SP is much more special then as a RP, and you don't see many relievers throwing in the high-80s/low-90s as a lot of SP do. That would adress MOST of the SP/RP dynamic problem, IMO. Would make more sense in the "RP are guys who can't start" reality for most pitchers.

Second, the Stamina rating could improve a lot. Today, it's a "as is" rating, rarely improves and never by a lot. And it's incredibly static inside a season. You can have a guy with high (say, 60s) stamina relieve the entire season. Then, if he is rested, have him start and go 8 innings. That rarely, if ever, happens in real life.

Stamina is built progressively through the season and the pitchers career. It should improve from the players 18-20yo to peak at mid/late 20s then drop at his mid/late 30s, depending on the player.

I say there should be a "maximum" stamina rating, similar to what we have now, and a "current" stamina rating, that changes more rapidly base on what a player has done lately, how much he has thrown, starting or relieving, injuries, et al. If a guy is in his 30s and has always been an 1-inning reliever, maybe his max stamina is high, but his current stamina is low and without much hope of improving further.

Third, SP and RP should be two different positions, i.e. like 2B and SS. So, if a guy has always been a starter, he has the stuff, the control, the pitches to relieve. But he never has, he isn't used to it, he would struggle at first. Like a SS who has the range, the arm and the quickness to play 2B, but never has in his professional life. With experience and exposure to the position, specially in training camp/minors, he would get used to it and be ok.

What do you think?

[Cinnamon J. Scudworth]

Originally Posted by Righty Groove

A lot of pitchers get relegated to the bullpen for lack of control, for too high of a walk rate. Getting a boost in control as a reliever would adress that.

Can you explain that a bit more? In real life terms, I don't understand how moving to the bullpen automatically gives a pitcher more control. It seems like it's mostly about giving him fewer chances to issue walks, not about lowering the walk rate (except perhaps as a function of a higher K rate, which would be modeled by stuff).

[Righty Groove]

Originally Posted by Cinnamon J. Scudworth

Can you explain that a bit more? In real life terms, I don't understand how moving to the bullpen automatically gives a pitcher more control. It seems like it's mostly about giving him fewer chances to issue walks, not about lowering the walk rate (except perhaps as a function of a higher K rate, which would be modeled by stuff).

You may be familiar with Tom Tango's work, right? Well, I won't cite him specificlly but, of the top of my head, he has a rule-of-thumb that, when a pitcher transitions to the bullpen, his K% improves 17% and his BB% decreases 17%, with the opposite happening when the transition is bullpen to rotation.

The reasons are difficult to prove or disprove, more anedoctal then anything. But, basically, relief pitchers throw more fastballs (higher in-zone%), hitters are less inclined to drive up his pitch count (as they would with a starter, perhaps), and when you focus on your best one or two pitches it's easier to throw strikes.

[Cinnamon J. Scudworth]

Thanks, that makes sense. I (obviously) haven't performed any kind of systematic observation of this in OOTP, but I'm just curious whether that already happens in OOTP to a degree, simply by virtue of the increased stuff rating.

[Craig Scarborough]

Originally Posted by Righty Groove

You may be familiar with Tom Tango's work, right? Well, I won't cite him specificlly but, of the top of my head, he has a rule-of-thumb that, when a pitcher transitions to the bullpen, his K% improves 17% and his BB% decreases 17%, with the opposite happening when the transition is bullpen to rotation.

The reasons are difficult to prove or disprove, more anedoctal then anything. But, basically, relief pitchers throw more fastballs (higher in-zone%), hitters are less inclined to drive up his pitch count (as they would with a starter, perhaps), and when you focus on your best one or two pitches it's easier to throw strikes.

Actually, what Tom Tango said was that K-rate when up 17%, but the walk rate remained the same. Pitchers don't dramatically increase their control because they're throwing "lights out."

From Tom's blog: "You can try slicing and dicing Rally’s data any way you like, but you will see little bias in eras. Basically, use the “rule of 17”: difference in BABIP is 17 points higher as starter. K/PA is 17% higher as reliever. And HR per contacted PA is 17% higher as starter. Walk rate is FLAT." http://www.insidethebook.com/ee/inde...ief_1953_2008/

Also - doesn't Markus already compensate for this? I thought I heard him mention that pitchers get a bonus when they are relieving as opposed to starting.

[kingcharlesxii]

Your second point is the biggest flaw in the current pitching system, I think. Stamina just doesn't work realistically right now - there needs to be a system in place to stretch out bullpen guys if you want them to start.

For the third one - I'd rather get rid of how the SP and RP labels work now. Instead, have them generated by the game - if a guy has the skills and is stretched out enough, label him SP-worthy. If not, he's a bullpen arm.

[Matt Arnold]

I agree with all of this.

I really like the pitching model as it is, but if we added this extra stamina adjustment, that'd make it even better. I would definitely like to see many more "borderline SP" type relievers, and have less guys who are simply dominant relievers. There's very few guys like your Esmil Rogers types, who take some time to get stretched out mid-season, and no way to do like what happened a few years ago with Wilson and Morrow, who both had to spend a good chunk of time adjusting to their new roles.

[olivertheorem]

I like the concept of incorporating time spent in each of the two roles into a SP/RP position rating.

I'd also like to get rid of MR/CL as positions and just call them all RP. Keep the MR/CL roles though.

[Cryomaniac]

Having an experience rating for Starting and Relieving, like there is for fielding positions, would be a good idea (although the fielding model could do with changing imho, but that's a different kettle of fish entirely)

[Fyrestorm3]

Originally Posted by Cryomaniac

Having an experience rating for Starting and Relieving, like there is for fielding positions, would be a good idea (although the fielding model could do with changing imho, but that's a different kettle of fish entirely)

That's not a half bad idea.

[injury log]

I agree that there are issues with the SP/RP model in OOTP (though it's a lot better now than it was five or six years ago). But I don't think there are any small adjustments that would make things appreciably better, and I don't agree with the OP about most of the issues he's identified.

We've long relied on concrete data to adjust OOTP modelling to make it more like real life. Markus has always been receptive to proposals that have data to back them up, but there is equally going to be reluctance to change things without any assurances that they improve the game.

Taking your points in turn:

Originally Posted by Righty Groove

First, most most MOST relievers are failed starting propects.

If by 'most', you mean 'the majority', then you're correct. But then that's true in OOTP as well; the majority of relievers are pitchers who started out with two good pitches and one bad one, or with decent enough Stamina to start, and who either didn't develop a third pitch, or who lost Stamina over time. They're starters in the low minors, and relievers in the Majors.

If by 'most' you instead mean 'almost all', that simply isn't true. I just scanned rosters in the AL Central, but all of these MLB relievers are pitchers with fewer than 10 starts in their entire minor and Major league careers (and most have 0 starts) :

Addison Reed
Jesse Crain
Ramon Troncoso
Chris Perez
Vinnie Pestano
Joe Smith
Cody Allen
Chen Lee
Bruce Rondon
Jose Valverde
Greg Holland
Tim Collins
Louis Coleman
Caleb Thielbar
Casey Fien
Josh Roenicke

That's just from five teams.

Originally Posted by Righty Groove

When "converting" a starter to reliever, not only should he get a boost in stuff, but also (and most importantly) in control and velocity.

In OOTP, most pitchers do get a boost to Stuff (and thus to their strikeout rate) when pitching in relief. This is partly because they rely on their best pitches in relief, and partly because they get a Velocity boost. That boost doesn't show up on the player profile, but it does show up if you play out games: a pitcher used in relief will throw slightly harder than if he is used as a starter.

There are reasons real life pitchers should exhibit better control when used in relief. For one, relievers more often have the benefit of facing same-handed batters, since relievers can be used situationally. Pitcher platoon splits generally can have quite a large walk rate differential. And if a pitcher converts from SP to RP, part of his control improvement can be accounted for by aging; as pitchers age, their control tends to get better. Finally, if you only study pitchers who convert from SP to RP and who pitch a decent sample of innings in relief, you're more than likely only studying pitchers who were reasonably successful in relief. So you have a sample bias issue: you're probably only looking at guys who, as relievers, have good enough control to pitch for a while.

Originally Posted by Righty Groove

Second, the Stamina rating could improve a lot. Today, it's a "as is" rating, rarely improves and never by a lot. And it's incredibly static inside a season. You can have a guy with high (say, 60s) stamina relieve the entire season. Then, if he is rested, have him start and go 8 innings. That rarely, if ever, happens in real life.

You would not believe the number of pages in the beta forum that have gone to discussing this very issue. If you can make a proposal which:

- does not involve much micromanagement for the user
- does not offer the user any exploits over the AI in developing starting pitchers
- does not seem completely arbitrary (by which I mean is not something like "if a pitcher doesn't start for two weeks, he loses 14 stamina points")
- has some support of real life data
- only allows pitchers to migrate from relief to starting with about the same frequency as occurs in real life
- enhances gameplay enough that it's worth the coding effort

then there is at least a modest chance it would be implemented. We've discussed many such proposals, and none were satisfactory for one reason or another.

To my point of view, coding a 'dynamic' Stamina rating would be a lot of work to model something that just isn't very important ('stretching out' pitchers). But of course what's important to me isn't necessarily what's important to others, so I won't press the point.

Originally Posted by Righty Groove

Stamina is built progressively through the season and the pitchers career. It should improve from the players 18-20yo to peak at mid/late 20s then drop at his mid/late 30s, depending on the player.

Do you have any concrete data to support this claim?

Originally Posted by Righty Groove

Third, SP and RP should be two different positions, i.e. like 2B and SS. So, if a guy has always been a starter, he has the stuff, the control, the pitches to relieve. But he never has, he isn't used to it, he would struggle at first.

Again, do you have any real life data which demonstrates this is actually what happens?

[kingcharlesxii]

Originally Posted by injury log

You would not believe the number of pages in the beta forum that have gone to discussing this very issue. If you can make a proposal which:

- does not involve much micromanagement for the user
- does not offer the user any exploits over the AI in developing starting pitchers
- does not seem completely arbitrary (by which I mean is not something like "if a pitcher doesn't start for two weeks, he loses 14 stamina points")
- has some support of real life data
- only allows pitchers to migrate from relief to starting with about the same frequency as occurs in real life
- enhances gameplay enough that it's worth the coding effort

then there is at least a modest chance it would be implemented. We've discussed many such proposals, and none were satisfactory for one reason or another.

To my point of view, coding a 'dynamic' Stamina rating would be a lot of work to model something that just isn't very important ('stretching out' pitchers). But of course what's important to me isn't necessarily what's important to others, so I won't press the point.

These are all valid concerns. Another idea I thought of is a starter "rust" factor. It seems like (in MLB) when a guy is switched from the bullpen to the rotation after not starting for a few months, they usually limit his pitches for two starts (generally around 70-80 for the first start and 80-90 for the second if they could throw 100-110 pitches normally). I think something as simple as adding the rust factor if a player hasn't started for two months that temporarily decreases the stamina could do the job pretty effectively.

I lack the skill to do the data collection but this is just first-hand observation from watching the Brewers this year, who did this multiple times due to injury/ineffectiveness of their rotation.

I think this is important because it gives the human player a big advantage if they can just pull a guy out of the bullpen and throw 130 pitches. They'll prioritize getting guys who can start in the bullpen where the AI doesn't seem to right now.

[Craig Scarborough]

Originally Posted by kingcharlesxii

These are all valid concerns. Another idea I thought of is a starter "rust" factor. It seems like (in MLB) when a guy is switched from the bullpen to the rotation after not starting for a few months, they usually limit his pitches for two starts (generally around 70-80 for the first start and 80-90 for the second if they could throw 100-110 pitches normally). I think something as simple as adding the rust factor if a player hasn't started for two months that temporarily decreases the stamina could do the job pretty effectively.

Yeah, it's called "stretching out" the pitcher's arm, and I hear about this all the time. I wouldn't mind a factor built in where a pitcher who has pitched in relief the majority of the year would have his pitch count as a starter have to be gradually increased as he gets used to his new role.

[kingcharlesxii]

Yeah that's the concept I was getting at. Markus might even be able to re-use some of the spring training rust code and have it just apply to stamina. Have it kick in if a player doesn't start for 6-8 weeks (I don't think I've ever seen a reliever get stretched out if they've started more recently than that). It would also have the side effect of making rehab starts for a pitcher with a long injury more useful. I haven't noticed a big drop-off plugging a guy in straight from the DL and I've rarely used the rehab assignment feature since it has been implemented because of that.

[HolyCow98]

I'm the same way with guys coming off the DL.

For me there are differences between my perception of how it should be and how it is in the game. I'm not saying I'm right and the game is wrong at all, just saying it feels off in certain situations, again just my perception.

As far as starting pitchers go, I know that if they have a very high stamina rating and if they have not thrown over the last 5 days he is going to be able to throw 110-125 pitches with no problem. It doesn't matter if it is opening day. It doesn't matter if he has been on the DL for 12 months. He will have the stamina to throw 110+ pitches.

These situations always feel "artificial" to me. They aren't too too a big deal for me but I do think they could be looked at. Perhaps a scaled back "rain delay" effect.