Justin Verlander: A Shot At 25?

Justin Verlander won his 20th game Saturday against the Minnesota Twins. Eric Seidman already covered his MVP candidacy this morning, and I have to agree that we should probably cool down on it a little bit. I’m sure my opinion surprises absolutely nobody — we here at FanGraphs can’t stand pitcher wins, and I’m no different.

However, as Eric mentioned, Verlander might have a shot at 25 victories this season. And as heartless and cold and robotic as I may seem on the outside, even I can recognize the weight of that accomplishment. It hasn’t been done since Bob Welch was led by the Oakland Athletics to a 27-win season in 1990 — and I choose these words for a reason. Welch wasn’t actually that great that year, posting a sub-2.0 K/BB ratio, a FIP above 4.00, and a 2.95 ERA which was only five points lower than the ERA which brought him a 17-win season the year before. But with a little help from his friends, Welch put together the winningest season we’ve seen in two decades. Verlander has 20 wins and only six starts (maybe only five) left, so he won’t catch Welch this year. But can he, realistically, get to 25 and put a name to the accomplishment that actually deserves it?

Simply going by the simple logic that Justin Verlander (or a comparably talented pitcher on a comparably good team) has a pretty good chance at winning every single time he goes out there, and he has more (or as many) starts remaining as wins needed, the answer is yes, he has a chance, even if it isn’t great. The numbers support this from a couple of different angles.

For our first estimate, let’s use something relatively simple. So far, Verlander has 20-5 win-loss record in 29 starts. We could use his .800 winning percentage to figure his chances, but we have to account for no-decisions, too. Verlander still has a “start win percentage” of .690, as he’s won 20 of his 29 starts this year. Were we to consider that his true talent rate for five starts, simple probability (five successes with a .690 success rate in five chances) gives us a 15.6% chance of him winning all five and ending with 25 wins. If he were to get six starts down the stretch, we get a whopping 29.1% chance of him winning five of the six and can tack on another 10.8% chance of him winning all six, for a grand total of 39.9%. Pretty great odds for such a rare feat.

However, projecting Verlander to continue at the same rate he’s been at this season is dangerous, as it assumes we know all we need to about not only Verlander’s pitching, but his team’s chance of winning and then his own chance at earning the decision from just 29 starts. The assumption at hand is simply too much.

I’ll present an alternative method here as well. Although it’s not perfect, it should show that Verlander, even despite the fact that he can’t count on the same combination of talent, performance, and teammate help that has gotten him this far, still has a shot at 25 wins nobody could have expected coming into the season. The method is based on the Marcel projections featured here — so simple a monkey could do it, as long as he had a working understanding of the SQL programming language.

For a Marcel-type projection, we weight the past three years on a 5-4-3 basis (most recent gets 5, then 4, then 3), and add in two seasons (68 starts) of an average pitcher. Remember, we’re concerned with start winning percentage here, not decision winning percentage. Of the 1,854 games started in the American League this year, 678 of them have been wins. That’s a start winning percentage of only .366 for the American League. That number goes with a .690 start win percentage this year in 29 starts, .545 in 2010 in 33 starts, and .543 in 2011 in 35 starts.

Performing the calculation, we get a .564 start win percentage — better than his recent years, but much closer to his career mark of .530 than his blazing mark of .690 this season. And that’s what we should expect. Verlander’s odds drop pretty dramatically with this calculation, but given the rarity of the 25-win season, his odds are still pretty high. If he gets just five starts, his odds drop to a mere 5.7%, but in the scenario with six starts, Verlander still has a 14.9% chance of 25 wins and a 3.2% chance at 26 wins for an 18.1% overall chance. I think if you told Tigers fans (or Justin Verlander, or Jim Leyland) that he would have a better than 1-in-6 chance at a 25-win season come August 29th, they would be pretty ecstatic.

These two methods, at the very least, give a pretty good range of probabilities for Verlander to reach the 25-win-mark. When we’re coming up with numbers with a floor of 15% and a ceiling (albeit an optimistic one) of 40% for a feat that hasn’t come around in two full decades, we know we have something worth watching for the rest of the season. And that’s always better than sitting around and debating who should win a hugely subjective and contentious award like the MVP.

Print This Post

Jack Moore's work can be seen at VICE Sports and anywhere else you're willing to pay him to write. Buy his e-book.

29 Responses to “Justin Verlander: A Shot At 25?”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Cloud Computer says:

    Very enjoyable write up – i hope he’s got a shot in his last start.

    Vote -1 Vote +1

  2. Griffin says:

    The only chance Justin has at 6 more starts is if they move him forward to pitch in the final series of the year (as it stands now, he would pitch the day after the final game). He won’t be moved up unless the division is in doubt on the final weekend of the year. I expect him to only make 5 more starts.

    Vote -1 Vote +1

    • CircleChange11 says:

      Yep no shot.

      Not when they likely want him to be as fresh as possible for the 1st round of the playoffs.

      Vote -1 Vote +1

      • Griffin says:

        That being said, he’s 8-0 in his last 8 starts, so it’s certainly not completely impossible to imagine him going 5-0 in the next 5 (especially considering the Tigers do not face a team with a winning record for the remainder of the season…well, the White Sox are technically a game over…but you get the idea)

        Vote -1 Vote +1

      • CircleChange11 says:

        Finishing the year 13-0 would be damn impressive.

        For years, I’ve watched Verlander pitch and wondered how anyone beats him. As his control/command continues to improve (or maintains current levels) it’ll continue to be tough.

        I know, in stats, you can’t look at a sequence of events … but rather look at each individual event. The probabilities for him to go 5-0 over his last 5 starts are long. Combine that with the 8 previous wins and now we’re looking at a 13-0 against all odds finish … even though in each of those 13 starts, he may have been a probable favorite to win the game.

        I lack the stats know-how to claculate the odds of him actually finishing the season on a 13-0 run.

        I would imagine over 34 starts, Verlander is projected to win about 29-30 of those (if you look at each game as a single event) … yet we all know even winning 2/3’s of your starts (even as an ace) is good.

        Vote -1 Vote +1

      • Griffin says:

        Winning is a binomial probability (meaning, if an event has a probability, p, of happening, then the probability of it happening twice is p to the 2 power, and in general p to the n power for n successive trials). Verlander has 20 wins in 29 starts. If we express the percentage of games where he has recorded a W as .6896, then we can take his winning percentage to the power of 5. This comes up with a number of approximately .156, or a 15.6 percent chance to win all of his remaining starts (this obviously does not take in to account things such as opponents, ect.)

        Vote -1 Vote +1

      • CircleChange11 says:

        That’s what I was referring to Griffin … lookijng at the strength of the oppossing hitters, as well as, the strength of their SP versus DET’s hitters, to come up with a “probability” for Verlander winning that game.

        Then repeating the porocess for the next game, and so on. Then, finally combining all of the probabilities to see the probability of winning all of the games consecutively.

        Vote -1 Vote +1

  3. glen garry says:

    I think Jeremy Guthrie has a better shot at 20 losses, which is even more rare and much more exciting.

    Vote -1 Vote +1

    • Detroit Michael says:

      I don’t know about the “more rare” claim. Losing 20 games has certainly been performed more recently than winning 25 games.

      Vote -1 Vote +1

    • Person says:

      Guthrie: 6-16, 4.42
      Lackey: 12-9, 5.98

      As we all well know, life is unfair. This is demonstrated not only in the above lines, but also that you could show some people the above lines and they would still take pitcher wins seriously.

      Vote -1 Vote +1

      • juan pierre's mustache says:

        im pretty sure even the most diehard defender of pitcher wins wouldn’t want to have anything to do with lackey at this point

        Vote -1 Vote +1

  4. Rex Manning Day says:

    For what it’s worth, assuming Detroit doesn’t shift Verlander’s schedule at all, these are his expected opponents in his last 5-6 starts:
    Royals (9/1-Home) – Team W%: .410
    Indians (9/6-Away) – .500
    White Sox (9/12-Away) – .504
    Athletics (9/17-Away) – .451
    Orioles (9/22-Home) – .405
    Indians (9/27-Home) – .500

    If they do move him around, he might have to face the Twins (currently sitting at .421), but otherwise he’d just play one of these five teams on a different day.

    His expected opponents have a combined Win% of .454. They have an average team wOBA of .313 (against an AL team average of .320). Verlander’s gone 9-2 with 0 NDs against these 5 teams this year, with a 2.88 ERA, 0.90 WHIP, 9.86 K/9, 2.00 BB/9, and 1.00 HR/9 over 81 IP. That’s roughly in line with his overall season numbers, albeit with a slightly higher ERA and K/9.

    Based on his previous outings, the White Sox probably represent the “biggest” threat to his 25 wins. They’ve hit him moderately well compared to the rest of the league this year (.225 BA compared to .189 for the league, with 6 HR), and Verlander does have a habit of giving up a bit more HRs away from home (plus, Chicago has the second highest HR factor in the Majors this year).

    Also, for what it’s worth, the As actually beat him up pretty good in his only start against them this year, wherein “beat him up” means 3 runs in 6 innings and only 6 Ks. Against 2011 Verlander, though, that’s actually pretty darn good.

    Anyway, small sample sizes blah blah blah, and also who knows how Detroit’s offense will perform against these teams, but it’s not like Verlander will be facing particularly stiff competition in these last starts. It’s still probably a long shot, but it’s at least a realistic long shot.

    Vote -1 Vote +1

    • buddy says:

      Well he’s already been pushed back to 9/2 for his next start.

      My guess would be:

      9/2 – Royals
      9/7 – Indians
      9/13 – White Sox
      9/18 – A’s
      9/24 – O’s

      If the division race is in doubt, then he’ll probably go on the 23rd and the 28th. Otherwise they’ll give him the extra days off.

      Vote -1 Vote +1

  5. John Kruk says:

    He could win 30 with the Yankees!

    Vote -1 Vote +1

    • jim says:

      given how he’s pitched this year, verlander could have gone undefeated with the yankees or red sox

      Vote -1 Vote +1

  6. JDub says:

    That upcoming sched looks easy but don’t forget he’s beaten the weakling Twins three times since the All Star break. .

    Vote -1 Vote +1

  7. Sultan of Schwwingg says:

    “we here at FanGraphs can’t stand pitcher wins”

    So, once they were misguided stats, but now they are even worse – they can’t be stood for. Looking forward to your next article assigning Young, Johnson and Mathewson as the best frauds of all time.

    You know, it would have been better to leave it as they don’t mean as much as the credit they’re given. No more was necessary. Today’s outright assault on pitcher wins by the SABER community, however, not only is shallow, in that they won’t consider them – they’re that useless, but also counter-productive to its being taken seriously by Main St America.

    Look at any pitcher WAR leader board for a given year and there they are, starters with the most wins. Funny how that works.

    Vote -1 Vote +1

    • Kyle says:

      Gotta agree to an extent, not every pitcher is winning games like John Lackey with the Red Sox due to the unstoppable lineups they can throw at opposing pitchers.

      But homeruns are a bit overrated considering most guys aren’t hitting them for thrilling come from behind wins to get their team into the playoffs. And saves, well… those are more overrated than wins. Most guys who have seasons where they win 20 to 25 games were studs. Occasionally a guy gets lucky due to a pitcher friendly park, but let’s see the guys who won 20 games.

      Gibson – 5 times
      Schilling – 3 times
      Feller – 6 times
      Carlton – 6 times
      Maddux – 2 times, won 19 games 5 times
      Koufax – 3 times
      Halladay -3 times, maybe 4 after this season
      Pedro – 2 times
      Nolan – 2 times
      Seaver – 5 times

      So yeah, Jose Lima or Matt Morris(while having a great season)might get 20 wins while Lincecum and Felix Hernandez might not ever see that number. But guys like Suppan, Silva, Pavano, and Liriano won’t either.

      Vote -1 Vote +1

      • Sultan of Schwwingg says:

        The irony is that SABR hates wins because they’re not perfect, since, annually, a handful of pitchers get undue credit for them. But neither is FIP or UZR or WAR perfect.

        Actually, on the whole I’d say those three are generally less perfect far more often than Wins.

        This mocking of wins is inappropriate. Outstanding season by JV.

        Vote -1 Vote +1

  8. grapesoda says:

    @ sultan:
    it’s not that hard to explain. of course a good pitcher has a better chance winning games than a bad pitcher. while it is not fair to say that wins mean absolutely nothing, we know that there are better ways to estimate a pitchers value than win-loss records.

    btw: i didn’t calculate this, but to me it looks like verlander has an even better shot at a more rare feature: the triple crown.

    Vote -1 Vote +1

    • Sultan of Schwwingg says:

      Yes, exactly. Good pitcher = “better chance winning games”. Bad pitcher, not so much.

      I’m pretty sure this is why the stat has endured for so long.

      Vote -1 Vote +1

  9. Atom says:

    The key to a 27 win season is simple.
    1 pretty good pitcher+
    A lineup of Rickey Henderson, Jose Canseco, Mark McGwire, Harold Baines, Dave Henderson
    1 bullpen of Dennis Eckersley (610 ERA+), Todd Burns (126 ERA+), Gene Nelson (239 ERA+), Rick Honeycutt (139 ERA+) and Joe Klink (184 ERA+).

    Vote -1 Vote +1

    • Kyle says:

      So explain Verlander’s 20 wins so far while Sabathia and Halladay don’t have as many even though they have much better lineups. Until recently Inge, and other guys hitting around .200 like Ordonez were starting daily for the Tigers. It’s not that easy… but wins while having been so overvalued are now undervalued.

      Obviously Welch winning 27 games was insane, cause he wasn’t even the best pitcher on that staff. Dave Stewart for around 4 years was their most clutch, dominant pitcher. In 1988, Orel Hershiser went 23-8 while completing 15 games(throwing 8 shutouts)with a weak offense of that had 99 homeruns, not one guy even close to 100 RBI, no .300 hitters and a team batting average of .248. He then destroyed the Mets and A’s(throwing 3 CG’s and having 2 shutouts.)

      I think that people are too hard on wins… obviously his next season he had an IDENTICAL SEASON but only went 15-15. His ERA jumped from 2.26 to 2.31, struck out more guys per 9, but only had 8 complete games and 4 shutouts. He’s a guy who could have easily had 45 more wins. Going 16-16 with a 3.06 ERA, 19-3 with a 2.03 ERA and 36 games, 11-8, with a 2.66 ERA with 45 games(20 starts)but 8 complete games.

      It’s just still a main facet of what you look for in your ace. What good is it to see guys leave early to let a weak bullpen blow the lead time and time again? Guys like Halladay, Verlander, even CC, or recently retired guys like Maddux, Schilling, the Big Unit, Nolan(he could have had so many more wins but got no run support time and time again), Gibson would fight if they were taken out of the game.

      I just think wins have gone from being overrated to now underrated. And with this change, I wonder if guys who don’t come close to 300 wins but have lower ERAs than their contemporaries considering their era in which they pitch will have a better shot at the HOF. Guys like Oswalt(who might not get 200 now), and guys I think have to get in like Smoltz, Schilling, and Tim Hudson if he continues for another 4 seasons at this pace.

      Vote -1 Vote +1

      • davisnc says:

        Tigers wOBA: .328 (104 wRC+)
        Phillies: .319 (97)

        That took five seconds. Much better lineup?

        Vote -1 Vote +1

  10. jacob says:

    “simple probability (five successes with a .690 success rate in five chances) gives us a 15.6% chance of him winning all five and ending with 25 wins. ”

    Is simple probability the same thing as a binomial distribution:


    Vote -1 Vote +1

  11. Mike B. says:

    A fun new stat for pitchers would be FIW, Fielding-Independent Wins (or Friend-Independent Wins). So we’d have the standard W-L numbers and then FIW + maybe Fielding-Independent Losses for a truer performance measure.

    Vote -1 Vote +1