Management by Baseball

In the last entry, I reasserted the critical success factor for successful Agile project staffing is having the balanced blend of skills on the team (that is, copy Baseball). But any project (and Agile ones in spades) is not static. If you've built the perfectly-balanced team, the right people with the right balance of skills & personalities, it's inevitable that as the project evolves/ rolls-on, that on any given day the team'll be unbalanced.

You already know that the ideal management tactics on one project will be different from the next, even with the same team. You probably already know that your management will have to adapt on a daily basis (though in practice, my observation of most managers of Agile and Lean operations suggests even if they know, most prefer a single static approach). Somebody is having a tough day with their family. Somebody who you count on, for example, for meticulous testing, is having a day she needs to do something open-ended or creative. To be Agile, you have to adapt how you manage most every day, even with the ideal team.

La Russa explains to Agile practicioners the core Baseball example of why and then what to do about it. The situation is a playoff, a best-of-five series against the Phillies, the best team in the league that season. Phils have won the first game, so it's almost mandatory to win the second game (lose the second game, you're 0-2 meaning you have to win three in a row against the best team in the league with their dominant pitching). The Cards have their best veteran starter on the mound, Chris Carpenter, and the Phils have one of the best post-season pitchers of his generation in Cliff Lee. Carpenter, on only three days rest instead of his normal four days, (abnormal for most starting pitchers and something he hadn't done in the previous ten seasons) struggled in both the 1st and 2nd innings and the Phils took a 4-0 lead -- a pretty ugly turn of events at a worst-possible time -- the time for the management team to do something to try to right the game/project. He explains in his most recent book, One Last Strike (p. 221):

Fortunately, (pitching coach Dave) Duncan was on the bench. He and Carpenter sat there between innings, and I knew that Dunc was talking with Carp about more than mechanics. As a pitching coach, you have to be part physiologist and part physicist to understand how the body works and and the ball moves through the air and part psychologist to understand how the pitcher's mind works.

Sound parallel to your Agile/Lean management requirements? It should.

I knew that he was helping Carp figure out a way to keep the Phillies right there, to block out whatever distractions there might have been about [...] his high pitch count, and the fact he was going on three days rest.

Carpenter goes on to hold the Phils in the 3rd inning with an eight-pitch, three-up three-down half inning and, because the Cards took a rally into his batting spot in the bottom of the 3rd inning, Carpenter was lifted for a pinch hitter. The Cards go on to win the game, the bullpen preventing the Phils from scoring another run. Two actionable lessons coming up...

After the game we learned from Carp that when he'd warmed up for the game he hadn't felt particularly strong (not surprising...shorter rest), so he went out there in the first couple of innings and tried to put extra effort behind every pitch. He was out of sorts mechanically. Carp told us this: if he ever got the change to pitch with three days' rest again, he [...] wouldn't try to force it. He would "pitch with what he had" and not try to add to it.

The two lessons Agile team co-ordinators need to take away from this are these. First, as Carpenter and his mentor La Russa learned again, in the maelstrom of a project falling short, we've all seen errors caused by passivity, but trying to close the entire gap by pushing "too hard" is a standard miscue to avoid, explicit in the XP playbook, but worth following in whatever Agile /Lean method you're using as a piece of your foundation.

Secondly, don't have MBWT (Management By Wishful Thinking) about what the team you have can do. At the juncture of schedule glitch/bump/crisis, it's critical to check in with all the individuals/players and figure out what you have so your expectations and designs are in line with what you can bring to bear on catching up.

PITCH WITH WHAT YOU HAVE & MANAGE WITH WHAT YOU HAVE
Just as in Baseball, even the most flexibly-staffed Agile team will have limited abilities to change its roster. When I'm spear-heading a project, I like to scout out talent with complementary skills outside the team and negotiate in advance the possibility of getting them on the team's roster temporarily. I've found this is a good way to emulate Baseball (where there's a set of minor league teams with marginal talent who can fill in the major league team's roster in a pinch). I call that The Bolt-On Talent gambit.

But even with bolt-on talent, most of the time, you're going to have to figure out how to help the team succeed with the team staffed as it is. La Russa and his coaching staff are master of this.

Starting on page 288 of the book, La Russa describes how his own mentors in a leading Agile think tank, the Chicago White Sox front office, modeled for La Russa how to handle a bullpen and starting staff. The two mentors were Paul Richards, one of the greatest American management theorists, and Ken Silvestri, the Chisox' organization-wide pitching coordinator.

They trained La Russa, then a beginning manager in a minors to have the young pitchers start but use older veterans to come into pressure situation to relieve. This fits the emotional algebra of pitching, because starters get many more chances to work through problems, but relievers have less time to the finish line, increasing the ambient pressure, and frequently, they are coming into a game because there's a "defect" (runners on base or a starter losing sharpness).

So LaRussa, like many experienced major league managers, prefers to have a bullpen staffed with experienced vets. But as he notes on page 288...

A veteran bullpen is the formula we've continued to use over the years -- except, ironically, in our last two World Championships, both of which were won with predominantly young relievers.

In September 2006, [veteran reliever] Jason Isringhausen had gotten hurt so our closer was Adam Wainwright, a rookie. Along with Adam, our other main right- handers in the penwere Josh Kinney, another rookie we'd called up, Braden Looper and Josh Hancock. Our two left-handers were Tyler Johnson, another rookie, and Randy Flores.

He goes on to describe one more "pitch with what you've got" moment during the late pennant drive in 2006 (a year they went on to win the World Series). They are stuck using Jeff Weaver in a key game. Weaver had been a promising young pitcher with some electrifying bright spots, but his inability to be consistent had pushed him to the margins late in his career, and he was absolutely not a starting pitcher you'd want to go to in a must-win game. In this game, pitching coach Dave Duncan decided to break the pattern (and therefore the scouting report the other team would use on Weaver), especially with left-handed batters who had been battering and frying Weaver all season.

Dunc, in his genius, told him "throw your curveball". And he curveballed them to death, mixing in just a few fastballs and changeups. He held them to two hits and no runs over five innings. But when he was ready to go through the lineup the third time, I yanked him. Again we went to our young bullpen - Johnson, Flores, Kinney, Wainwright. We won that {key] game 2-0.

HEWING TO STANDARDS & BREAKING THEM, TOO
So while LaRussa and Duncan went against standards with a flaky pitcher and throwing an unprecedented set of tactics, they had to go with what they had. OTOH, they stuck to standards to protect the win.

Over 95% of starting pitchers are markedly less effective the third time through the opposition's lineup than they are the first two times through. Batters get to see what the pitcher is doing, what's working well for him and not, and can time and better judge what's approaching the plate. Weaver's curveballing the opposition "to death" was a surprise, delivering a marginal advantage to an otherwise shaky pitcher, destined to be a non-surprise the third time the batters got to experience it.

Ergo, even while throwing away a piece of standard operating procedure, LaRussa and Duncan embraced the third time through standard. They didn't try to ride a cute trick past its effectiveness, preferring to bank the advantage.

BEYOND BASEBALL INTO AGILE
As a Coach or Scrum Master or [insert title here] striving to maximize team effectiveness, it's essential you follow the Baseball tenet of "Pitch with what you got", shaping your designs and tactics to maximize each talent's pattern to result in the most winning outcomes. At the same time, you have to be Baseball, and blend the known- winner patterns with experiments and norm-breaking designs that are more apt to succeed than the standby patterns when the situation calls for it.

Be more like LaRussa. Be more like Baseball. That is, Win.