The primary metrics by which coaches and managers in professional sports are measured are wins and losses. Wins and losses exist for players, particularly for baseball starting pitchers, but there are more descriptive statistics which are in the mainstream (earned-run average, to use the most famous example). For coaches, one’s record is what we have, so it’s largely what we use.

Popular perception of St. Louis Cardinals manager Mike Matheny has largely fluctuated with the performance of his team. In his first four seasons as manager, Matheny and the Cardinals reached the postseason; the run included a trip to the World Series and a 100-win regular season. Matheny had more than his fair share of critics in more analytically-driven circles but those who supported him (or at least didn’t actively dislike him) could point to his win total. But in the next two seasons, the Cardinals missed the playoffs twice, and in 2018, the team is teetering on the brink of the second Wild Card position.

This decline in performance for the Cardinals probably has little to do with Matheny’s management–it’s a reflection of several factors, most notably the talent which comprises the roster. But criticism of Mike Matheny has increased for largely unfair reasons–if anything, Matheny’s tactical ability has improved, even if too slowly for my tastes. In general, of course, fans should want a manager who is the best possible manager, but sometimes whether or not he is successful does not reflect whether or not he is good.

On June 17, the Cardinals defeated the Chicago Cubs 5-0 to salvage some value from what otherwise could have been a disastrous weekend set against their chief NL Central rivals. The Cardinals won the game largely on the strength of their pitching–starter Jack Flaherty went for five shutout innings, while relievers John Brebbia and Austin Gomber supplemented this with scoreless frames of their own. And following the seventh inning, in which a solo Matt Carpenter home run gave the Cardinals a three-run advantage, the Cardinals’ odds of winning the game, according to Baseball Reference’s measure of Win Probability, stood at 96%.

It may be difficult to imagine what 96% is if you have not seen Win Probability regularly, so for some perspective, here are a few notable Win Probabilities in St. Louis Cardinals history.

• When Jack Clark belted a home run to flip the result of Game 6 of the 1985 NLCS against the Los Angeles Dodgers, the Dodgers’ Win Probability was 81%.
• When David Freese came to the plate in the bottom of the 9th inning against the Texas Rangers in Game 6 of the 2011 World Series, the Rangers’ Win Probability was 92%.
• When Daniel Descalso came to the plate with a two-run deficit and two outs in the top of the ninth inning in Game 5 of the 2012 NLDS, the Win Probability for the Washington Nationals was 86%.

These are all-time, legendary, dramatic comebacks for the St. Louis Cardinals. And the probability of a Chicago Cubs comeback on Sunday was lower. A three-run lead, particularly a three-run lead with two innings of baseball remaining, seems less than completely safe, but relatively speaking, the Cardinals were a lock to win. And yet the Cardinals opted for Jordan Hicks, the team’s best reliever over the preceding month-and-a-half, to pitch in the eighth inning.

As Hicks had pitched the previous day, he was unavailable to pitch on Monday night. On Monday, with Hicks unavailable and closer Bud Norris already out of the game having pitched the ninth inning (and with Francisco Pena having pinch-hit for him in the top of the tenth), the Cardinals turned to Matt Bowman. The Cardinals’ Win Probability entering the inning stood at 81%. And Bowman, who has been sub-Replacement Level on the season, allowed two runs and the Cardinals lost the game.

It is entirely possible that the result Monday night would have been the same even had a rested Jordan Hicks been available to pitch, but it would be difficult to fairly blame Mike Matheny had the team simply lost with the best available personnel not performing up to its potential (which isn’t to say that Matheny wouldn’t have been criticized, but it would have been unwarranted). But in all likelihood, the Cardinals lost on Monday night because of Sunday.

A second example of the Cardinals utilizing one of their more reliable bullpen options in a low-leverage situation happened on Sunday night in the top of the ninth inning, when Bud Norris pitched with a five-run lead; by this point, the Cardinals’ 96% Win Probability had jumped to well above 99%. On paper, pitching Norris was even less defensible than pitching Hicks, but unlike Hicks, Bud Norris hadn’t pitched in six days, so there was an argument to be made that he needed work to remain fresh.

But if the Cardinals (correctly) were compelled to preserve Jordan Hicks but also wanted to give Bud Norris some work, there was a very simple solution to this–let Norris pitch the eighth inning. After all, the batters coming to the plate in the eighth were Kris Bryant, Anthony Rizzo, and Willson Contreras, arguably the three most dangerous hitters in the Cubs lineup. If the Cardinals were planning on using Bud Norris and (fill in name of lesser, non-Jordan Hicks pitcher), why not use Norris to face the stronger competition? Wouldn’t this lead to a (marginal) increase in your likelihood of winning the game?

The probable answer is the save rule. It is probable that setting up Bud Norris to get a save (when Norris began warming up, it was a 3-0 game and therefore, had the Cardinals not scored in the bottom of the eighth, it would have been a save situation) figured into Mike Matheny’s mind, but it is also likely that decades of managers, including hundreds that are not Matheny, managing with the save in mind has affected the perception of a three-run lead.

The save rule indicates that if a closer enters a game in the ninth inning with a three run lead and he records the next three outs, he earns the save, despite the fact that the game is pretty well locked down. And yet if the lead expands from three runs to four, despite a minor change in Win Probability, it dramatically affects pitcher usage. On Sunday, ESPN color commentator Alex Rodriguez noted that the only reason Norris would still pitch the ninth inning was because he had already warmed up, and this is probably true. He could have pitched in the eighth inning, against the heart of the Cubs lineup, and it would have been justifiable on the grounds that he needed to pitch, but he wouldn’t have gotten the save.

Pitching Jordan Hicks on back-to-back days in a game that the Cardinals already have a 96% chance of winning is like using a gun to turn out the lights at your house–yes, it’s probably going to work, but pretty much any solution was probably going to work–this solution just happens to have a bunch of really obvious downsides to it. This was a strategy once used by Homer Simpson. Generally speaking, you don’t want your baseball team’s manager to be compared to Homer Simpson.

Have the relative struggles of the Cardinals bullpen this season perhaps made the realistic odds of a victory on Sunday less than 96%? Sure. But as bad as the bullpen has been, it has not been “allow three runs in two innings” bad. And if it were that bad, then caring about game results as they relate to the standings is pointless, because a team with a 13.50 bullpen ERA would finish a couple dozen games out of a playoff spot.

And yet because the Cardinals won on Sunday (in a shutout! Against the team’s biggest rival! To end a frustrating losing streak!), Mike Matheny was able to avoid criticism of what the numbers say was managerial malpractice. But the decision to pitch Jordan Hicks, ignored as it might have been at the time, was the worst managerial decision in a crushing loss on Monday which included intentionally walking the game-winning run because, to paraphrase Matheny himself, you gotta set up a double play. For a man who purports to be an enormous chess aficionado, Mike Matheny showed a shocking inability to look multiple moves ahead.