Phillies vs Rockies, Game #4 Simulation

by Xeifrank on Oct 12, 2009 12:48 AM MDT Comment 17 comments

I used my simulator to simulate 100000 baseball games between the Phillies and Rockies using what to the best of my knowledge will be close to today's starting lineups. The simulator outputs a win probability for each team, along with the average total runs scored, a distribution matrix of all the final scores and how often they occurred, and an averaged box score tally of all games combined. You may find it strange that there are so many one run games listed as the most likely score, and that the home team is always favored to win in the top couple of scores. This is due to basic math and the way that baseball rules play out of having the home team bat last. Here is a great article explaining this phenomenon.

Today's Results... (Last simulation ran at 1030PM PDT, Sunday)

Visitors	Home	Pitching Matchup	Favorite	Vegas Win Prob	Simulator Win Prob	AccuScore
PHI	COL	C.Lee vs U.Jimenez	COL	51.69%	48.41%	46%

Skinny: Simulator Fun Facts... My simulator has the Phillies as very slight favorites to wrap up this series and move on to face the Los Angeles Dodgers. Vegas is waiting to put out their line (sissy) the day of the game. My hunch is that they will have the Phillies favored by 4 or 5 percent more than the sim. AccuScore will likely have a similar prediction as the sim. Let's wait n see. Results are in for Vegas and AccuScore. Vegas has the Rockies as ever so slight favorites, about the same margin that the sim has the Phillies as favorites. AccuScore comes in with the Phillies as a more moderate favorite, upping their win probability to 54%. My guess is that the betting public is factoring in the irrational of the Rockies having their back up against the wall.

Top 50 Most Likely Scores
1	COL 3-2
2	COL 2-1
3	COL 4-3
4	PHI 3-2
5	COL 5-4
6	PHI 4-3
7	PHI 2-1
8	PHI 4-2
9	PHI 3-1
10	COL 3-1
11	COL 4-2
12	COL 1-0
13	COL 6-5
14	PHI 4-1
15	PHI 5-2
16	PHI 5-4
17	PHI 5-3
18	COL 4-1
19	COL 5-2
20	PHI 5-1
21	COL 5-3
22	COL 2-0
23	PHI 6-2
24	PHI 6-4
25	PHI 2-0
26	PHI 6-3
27	PHI 6-5
28	COL 3-0
29	COL 5-1
30	PHI 1-0
31	PHI 3-0
32	PHI 6-1
33	PHI 4-0
34	COL 6-3
35	COL 7-6
36	COL 6-2
37	COL 4-0
38	COL 6-4
39	PHI 7-2
40	PHI 7-3
41	PHI 7-4
42	PHI 5-0
43	PHI 7-1
44	PHI 7-5
45	COL 6-1
46	PHI 6-0
47	COL 7-3
48	PHI 8-2
49	PHI 7-6
50	COL 5-0

Game Pitching Results
Pitcher	IP	SO	BB	HR	WHIP	FIP
C.Lee	7.0	5.873	1.227	0.428	1.222	2.844
U.Jimenez	6.7	6.566	2.695	0.498	1.297	3.413

Note: Keep in mind I did this simulation before knowing the actual starting lineups, so some of the players I used may not be starting. Many of the "Players Most Likely To" stats depend on having the correct lineup. The lineups I used are listed below. Chances are you won't see a big difference in win probability from having the lineup order slightly off. Picking the correct starters is more important, especially if one of the better players is taking the day off. But you still won't see a big change in win probability if the starter and backup are interchangeable.

Simulation Lineups
	Name	wOBA	Name	wOBA
1	J.Rollins	.3226	C.Gonzalez	.2779
2	S.Victorino	.3373	D.Fowler	.3279
3	C.Utley	.3559	T.Helton	.3459
4	R.Howard	.3485	T.Tulowitzki	.3559
5	J.Werth	.3241	G.Atkins	.3275
6	R.Ibanez	.3437	Y.Torrealba	.3067
7	P.Feliz	.2819	R.Spillborghs	.3353
8	C.Ruiz	.2996	C.Barmes	.3104
9	C.Lee	.1364	U.Jimenez	.1609

Eat. Drink. Be Merry. But the above FanPost does not necessarily reflect the attitudes, opinions, or views of Purple Row's staff (unless, of course, it's written by the staff [and even then, it still might not]).

0 recs | Comment 17 comments

Story-email Email Printer Print

Comments

I'll take any of those outcomes

that result in a Rockies win.

by avsfan4ever33 on Oct 12, 2009 8:39 AM MDT reply actions 0 recs

Is CarGon's wOBA

hurt by the fact that he hasn’t played a whole lot this season?

I don’t recall exactly how wOBA is calculated.

Posting 65 comments/day since June '07.

by DbacksSkins on Oct 12, 2009 11:43 AM MDT reply actions 0 recs

wOBA is a rate stat

so no, it’ll just reflect the level of play.

WAR and wRAA will be hurt by playing time.

FREE CHRIS IANNETTA

At least against LHP, I mean COME ON

by Andrew Martin on Oct 12, 2009 11:51 AM MDT up reply actions 0 recs

Okay, that's what I thought,

but I wasn’t sure how it was weighted.

CarGon might conceivably be hurt by his youth and the small sample size, however. That wOBA just seems awfully low for him, and it certainly FEELS like he’s played better than that. At least late in the season.

I have a shrine dedicated to Mark Reynolds, wherein I keep his bobblehead, signed baseball, and jersey T-shirt.

by DbacksSkins on Oct 12, 2009 12:00 PM MDT up reply actions 0 recs

wow, that wOBA is freaking godawful

I don’t think you can really use a 3 year weighted with guys as young as Gonzalez, esp making a transition from Oakland to Colorado

FREE CHRIS IANNETTA

At least against LHP, I mean COME ON

by Andrew Martin on Oct 12, 2009 12:26 PM MDT up reply actions 0 recs

Yeah,

I have a bit of trouble seeing CarGon with a wOBA below Torrealba, Ruiz, Feliz, and Clint Freaking Barmes.

I have a shrine dedicated to Mark Reynolds, wherein I keep his bobblehead, signed baseball, and jersey T-shirt.

by DbacksSkins on Oct 12, 2009 12:47 PM MDT up reply actions 0 recs

OTOH,

the simulation has Toolowitzki hitting as well as Utley in this game.

I have a shrine dedicated to Mark Reynolds, wherein I keep his bobblehead, signed baseball, and jersey T-shirt.

by DbacksSkins on Oct 12, 2009 12:50 PM MDT up reply actions 0 recs

Take a look at CarGon’s wOBA over the past two years. Take a look at Barmes’ wOBA over the past two years. Not much difference. Now adjust for lefty/righty splits.
vr, Xei

by Xeifrank on Oct 12, 2009 12:50 PM MDT up reply actions 0 recs

If you think only the past 2-3 months of data should be looked at when making a hitters or pitchers projection then you may have an argument against the sims output or that the most recent data should be weighted “extremely” heavier than the previous data, then you also may have an argument. If you do that, you may improve one players projection but worsen ten others.
vr, Xei

by Xeifrank on Oct 12, 2009 12:55 PM MDT up reply actions 0 recs

No, after you explained how you got the numbers,

it made a lot more sense. It was just striking to see him that low before you clarified it.

I have a shrine dedicated to Mark Reynolds, wherein I keep his bobblehead, signed baseball, and jersey T-shirt.

by DbacksSkins on Oct 12, 2009 1:06 PM MDT up reply actions 0 recs

But the concept of What have you done for me lately is a factor to consider

Tango pointed out that hot hitters continue to be hot, while cold hitters hit should be expected to hit at their career wOBA.

that said, I really don’t buy much into projections for young players; they are simply too volatile to really figure out.

FREE CHRIS IANNETTA

At least against LHP, I mean COME ON

by Andrew Martin on Oct 12, 2009 1:18 PM MDT up reply actions 0 recs

Well, it’s hurt by a couple of things in the sim.

1) LHP vs LHB (.602 OPS vs LHP)
2) Cliff Lee is a very good pitcher.
3) Input projections are multi-year weighted. Meaning pre-2009 stats go into the hitter projection formula. Of course the more recent stats are more heavily weighted, but this can have some impact as his 2008 numbers aren’t nearly as good as his 2009 numbers.
4) Since his sample size of career ABs are fairly low, he has more regression to the mean than a more established player may have.

Keep in mind the wOBA printed in the FanPost is not his “input” wOBA, it is the output. For example if CarGo has a (just making up these numbers for an example) wOBA of .355 vs the league and that is his “input” projection to the sim. Obviously, this number will get smaller facing a left handed pitcher of the quality of Cliff Lee. Thus the .2779 number.

Thanks for the thoughtful question. :)

vr, Xeifrank

by Xeifrank on Oct 12, 2009 12:00 PM MDT up reply actions 0 recs

Ahhhhh.

Now that makes a lot more sense. Thanks for the clarification.

I have a shrine dedicated to Mark Reynolds, wherein I keep his bobblehead, signed baseball, and jersey T-shirt.

by DbacksSkins on Oct 12, 2009 12:03 PM MDT up reply actions 0 recs

but that's a pretty big chop off for that

are you factoring in Lee’s wOBAa as well or what

FREE CHRIS IANNETTA

At least against LHP, I mean COME ON

by Andrew Martin on Oct 12, 2009 12:26 PM MDT up reply actions 0 recs

wOBA is a by product of all the other output stats. wOBA is not an input projection into the simulator. Other things like K%, 2B%, 3B%, HR%, BB%, BABIP, Speed Score are for hitters. For pitchers, it’s K%, BB%, HR% etc… Of course there are more things that adjust these input numbers like park factors, split history, HFA etc… The wOBA printed in the FanPost is the output of Carlos Gonzalez’s results every time he comes up to bat in the simulator. So it’s not like I enter Cliff Lee’s wOBA against and Carlos Gonzalez’s wOBA and then come up with a likely wOBA. It is calculated from the wOBA formula, and based on the output of all the head to head matchups from Carlos Gonzalez vs Cliff Lee and the Phillies bullpen. Hope that makes sense. It’s not easy to explain without going into a lot of details.

vr, Xei

by Xeifrank on Oct 12, 2009 12:48 PM MDT up reply actions 0 recs

gotcha, I didn't realize you were using the component inputs

FREE CHRIS IANNETTA

At least against LHP, I mean COME ON

by Andrew Martin on Oct 12, 2009 1:16 PM MDT up reply actions 0 recs

Thanks for all the constructive peer review. I appreciate it very much. The CarGon projections will be something I will look into in the offseason as a possible area of improvement. They may be correct as is, but I will atleast look into it closely.
Thanks! And good luck tonight. I hope the umps don’t have a big influence on tonight’s game.
vr, Xei

by Xeifrank on Oct 12, 2009 2:37 PM MDT reply actions 0 recs