2024 MLB First 5 Experiment

  • Start date
  • Replies 82 Comments
  • Views 688 Views

Matty

0-fers Will Happen™
Since
Jan 26, 2010
Messages
50,333
Score
5,825
Tokens
0
Ok, so thanks to Rogatien's suggestion in the other thread we're going to test first 5 innings plays for the next month. The implementation is pretty straightforward - instead of weighing the pitching 60/40 between starting and relief pitching, we're going with 100% weighing on the starting pitcher. This isn't perfect obviously, as plenty of starters end up getting pulled before the end of the 5th, but that's why I built a bit of a buffer between what the model spits out and what I'm willing to play (i.e. I wouldn't play a 2-cent theoretical edge).

I am betting these, but since I'm simply testing a new idea, I will keep a separate record for it. I'd say don't bet these as I am likely to tweak the formula over the coming days/weeks.

1st 5 Innings Test Day 1 - Monday April 29
STL -108
WSH +120
TBR -114
MIN -150
PHI -138
LAD -120
ATL -138
 
Adding

TOR -150
 
Does this first 5 model factor in when teams use an opener pitcher vs the traditional starting pitcher.
 
I just pull the pitcher's WAR value then tweak the team's winrate as if the guy was their only pitcher. It does mean that the model likely underrates relief or opener pitchers that get used as starters. WAR doesn't think much of short relief pitchers.

This is not an exact science. :handshake:
 
Does this first 5 model factor in when teams use an opener pitcher vs the traditional starting pitcher.
Out of curiosity, I checked yesterday's games and LAD/TOR is one where I probably should've taken into account the fact that Grove probably wasn't gonna pitch more than a few innings. But after replacing him with Yarbrough (whose WAR is 0.8 lower than Grove's), the effect on the modeled ML is just 2 cents. It's largely inconsequential.

It would be a bigger deal IF a team used a mediocre opener followed by an ace starter, AND their batting WAR was depleted due to injuries to star hitters. It probably doesn't happen enough to bother. Miami and ChiSox are currently the two teams that are super weak in batting, where such a scenario would skew the model's output significantly.
 
I feel like you will have to have a very firm grasp of when a team is using an opener in order for this to give you the edge you're looking for, no? But I suppose this can be said for full game plays as well.

The more banged up a teams starting rotation is, the more likely the will have an opener. The only team off of the top of my head who use openers without a banged up staff is the Rays (who I believe "invented" the opener).

I'm not even sure if the Rays still do it semi-regularly, though.
 
I feel like you will have to have a very firm grasp of when a team is using an opener in order for this to give you the edge you're looking for, no? But I suppose this can be said for full game plays as well.
I don't think so. I'll play around with it but I don't think it matters much in the grand scheme of things.
 
I don't think so. I'll play around with it but I don't think it matters much in the grand scheme of things.
What's your logic behind it not mattering? (Not being confrontational, genuinely curious.) I feel like weighing starter 100%/bullpen 0% for a 5 inning stretch that will by design be 60-80% bullpen seems misaligned.
 
The short answer is that the vast majority of bullpen pitchers are around 0 WAR. The very best are around 2.

I'm not saying it doesn't matter, as much as I'm saying that I have no real way to correct for it, and that playing the occasional neutral EV bet is not a big deal.
 
(FYI the main reason I'm betting these without being certain of the method, is that I'm trying to satisfy a bonus rollover requirement at TheScore Bet. I need to rollover $5k by May 23 to get the max bonus.)
 
I'm glad it spat out more plays than the full
 
Five innings is not a lot of innings.
 
Feels like I'm lucksacking this shit so far. 3 wins 1 push.

:woohoo:
 
1st 5 Innings Test Day 1 - Monday April 29
STL -108
WSH +120 :greencheck:
TBR -114 :greencheck:
MIN -150 🟡
PHI -138 :greencheck:
LAD -120 :greencheck:
ATL -138 🟡

Adding

TOR -150 :greencheck:
YTD 5-0-2, +5.2u
1000030250.gif

Going to bed on a high note, nighty night Gamelive :woohoo:
 
Last edited:
yes :yesss:

I'm expecting no less than +69 units season
 
Weeeeee

We got lucky af last night. Weird thing about betting first 5 is that there is no reason for either team to try and win that specific race to 5.

But yeah, good start either way, and first 5 prices seem a lot softer than full-game lines. They vary a lot between my 3 books.
 
YTD 5-0-2, +5.2u

Tuesday April 30


COL +115 (originally posted/played in the Polaroid thread)
NYM -128
SFG -135
MIL -170
SDP -165
SEA -110
ARI -115
PIT -130

:letsplayball:
 
Last edited:
YTD 5-0-2, +5.2u

Tuesday April 30


COL +115 (originally posted/played in the Polaroid thread)
It's a BING BONG