The Rockies still want a bonanza in return for Troy Tulowitzki. Now 30, he’s under contract for six more years at $20 million per, and he had hip surgery in August. When healthy, Tulo is a premier talent, earning his salary and then some. But he’s been hurt a lot: In eight years since becoming a regular (age 22-29), Tulo averaged 117 games, missing 28% of the schedule to injury.

Yet, he’s done enough to rank 10th in total bWAR for that span (38.0), and a runaway 1st among shortstops. While injuries cost him nearly half the last three seasons, his rate stats reached new highs.

Optimists could note that Tulowitzki has bounced back from each prior injury:

• In 2008, his second full year, two separate DL trips cost him 60 games. But he finished that season on fire, and played like an MVP in 2009.
• In 2010, a pitch broke his wrist and cost him six weeks in mid-season. But he hit .323/1.020 afterwards, earning September’s Player of the Month with 15 HRs. 2011 was another great year.
• In 2012, a late-May groin tear needed surgery that knocked him out for the rest of the year. But he was back to peak form in 2013 …
• … right up until he cracked a rib on a defensive dive, missing a month in mid-year. He slumped for a while afterwards, but again finished strong, hitting .326/.923 in the last 40 games.
• 2014 could have been his best yet, but he tore a labrum running out a ground ball just after the All-Star break. His 5.5 WAR was the 4th-best ever while playing less than 100 games.

Some of these injuries don’t affect his forecast, but groin and hip surgeries are a concern, especially for a shortstop. Over the last eight years, Tulo ranks 4th in defensive WAR, but 31st in WAR batting runs. If he can’t stay at short, or can’t play it as well, his value takes a big hit.

This all makes it hard to gauge his worth. SBNation’s Grant Brisbee explored what Tulo might command as a free agent: Probably more total dollars than his current contract, even if the risk holds down the guaranteed years to five or six. But money alone won’t land him.

And the odds seem against him playing like an All-Star for the next five years. I studied this backward and forward, using data from Baseball-Reference.com:

__________

First, I took the 141 players with 20+ WAR during age 30-34 — a star-level sum reached by 24 players of all ages in the last five years — and looked backward at their games played from age 27-29.

Of those age 30-34 stars, just one played fewer games than Tulo’s 264 during 27-29, and all but these three played at least 50 games more, except those who were in the service, in the minors, or full-time pitchers for some or all of age 27-29:

• Paul Molitor, 258 games at 27-29, then averaged 138 G and 5.1 WAR for 30-34.
• Augie Galan, 298 games at 27-29, then averaged 122 G and 4.1 WAR for 30-34.
• Tony Phillips, 308 games at 27-29, then averaged 150 G and 4.6 WAR for 30-34.

Those three had injury records on a par with Tulo, so the path to a high-level recovery is not untrodden. But suppose we raise the bar to 27 WAR for age 30-34 — still less than Tulo’s healthy rate, and just enough to crack the top ten for 2010-14. All but three who did that played at least 100 more games than Tulo at 27-29. Adjusting for schedule length cuts the exceptions to one:

• Jim Edmonds, 342 games at 27-29 (78 more games than Tulo), then averaged 147 G and 6.4 WAR for 30-34.

That’s a slim hook on which to hang an empty-the-farm-system trade.

__________

Second, a forward-looking study focused on injury at age 28 or 29.

From 1919-2009, I took those who had 20+ WAR through age 27, and played into their 30s. I removed any whose wartime service would affect the study. Then I separated those who missed 40+ games in a season at 28 or 29, not due to service or holdout. This left 50 players in the injured group, and 137 others.

I compared the groups’ games played in the periods through-27 and 30-34. To make sure that the latter period didn’t merely reflect lingering effects of the 28/29 injury from which the player later recovered, I also measured them at age 32-34, omitting at least two seasons after the injury.

The groups’ average games were almost identical through age 27, but not later on:

Those injured at 28 or 29 played 13% fewer games than the uninjured group during both later periods.

__________

Some think that Tulo’s high production per game means that his tendency to miss 40+ games a year isn’t a big deal, as long as there’s a decent backup. I don’t think a general manager would agree.

First of all, GMs would much rather build around guys who have proven durable. The 12 position players who earned $20 million last year averaged 151 games in the three years before those deals were signed, and only one seemed a real injury risk. Second, a team that would bid for Tulo hopes to make the postseason; but his year ended at least two months early in two of the last three campaigns. The fear of him missing October would be a real factor in a GM’s thinking.

Third, most teams don’t have a backup shortstop they’d be comfortable starting for 40+ games, so getting one is another expense. The Rockies, who know best the risks involved, tried to get by with cheap backups for Tulo the last three years. Those backups played below “replacement level,” so although Tulo averaged 6.8 WAR per 162 games, the shortstop position averaged less than half that.

But the worst thing about that viewpoint is ignoring how aging affects a shortstop’s value. I studied the 57 shortstops through 2009 who had at least 15 WAR during age 25-29 and played into their 30s. There were no military disruptions in the study pool. During age 30-34:

• OPS+ fell by just 7%, and
• Games played fell by a modest 14%, but
• Total WAR plunged 38% (from 21.8 to 13.5 WAR).
  That’s much more than suggested by the drops in batting and playing time, mainly because
• Defensive WAR fell 50%.

(I also ran numbers for the group’s better half according to total WAR, oWAR, dWAR, and WAR/162. In each case, total WAR dropped 38-40% at age 30-34, and the defensive/positional value took a greater hit than offensive performance.)

Much of the steep drop in defensive value came from changing positions: At 30-34, one-third of these guys mainly played elsewhere than shortstop. Nine moved to 3B, two to 2B, two to 1B, one to CF, and three had no primary position. By age 34, only 24 were still regular shortstops.

So, Tulo’s at the age where shortstops often change positions — especially tall shortstops. Tulo is listed at 6′ 3″ (75 inches). Drawing the line one inch lower, I counted those with 500+ games at shortstop in two age ranges:

(And one of those tall, over-30 shortstops was Derek Jeter, so….)

The chance that Tulo might have to move is an added complication for a team that pursues him. Shuffling value around the diamond isn’t easy, now that trades are more complex than ever.

Getting back to the WAR-per-game issue … Tulo at 25-29 averaged 6.9 WAR/162 (4.8 WAR in 106 G per year). But if that rate falls at the 28% group average for 30-34:

• Even a rise in playing time to 120 G/year would produce only 3.7 WAR per year.
• With no change in playing time, 3.3 WAR/year.
• With the group’s average drop in games, 2.8 WAR/year.

__________

One last angle: Eighteen shortstops made 20+ WAR at age 30-34. Only two of them missed 50+ games to injury any year in their 20s, after getting established: Bill Dahlen once, Barry Larkin twice. None missed 70 games in a year in their 20s. Tulo had three years missing 60+ games, two missing 70+. The weight of history is against him joining this group.

__________

There are always outliers. As one who delights in a daily box-score hunt for the “never-before,” I would be a hypocrite to act as if precedent utterly guides anyone’s future. Tulowitzki could turn out like Molitor, Larkin, Joe Cronin and Max Carey, all significantly injured at 29, but with many fine years ahead.

I just don’t want my team to place that bet.