- 07/10/2012, 11:01AM ET
MagicSpecs said 07/10, 11:09 AM
I propose that the next inequality in MLB to be exploited to gain an edge is more effective injury management, with emphasis on prevention and rehabilitation.
To win games you need good players, and you need them to play. Having an All-Star team at your disposal is no good if they're all on the DL ??? you wouldn't even beat the Astros (maybe). Losing just one key player to injury could break your season ??? see Dodgers, LA: 24-11 before Matt Kemp first hit the DL this season, 23-29 since.
Injuries are at epidemic proportions in baseball.
By my count there are 22 players in MLB playing in $100m+ contracts in 2012. Two of them (Crawford and Howard) currently combine for a whole 2 games played in '12. Another 7 of our $100m men have seen the DL this season. Of the remaining 15, 8 took a DL trip in '11.
Only 8 of the 22 current $100m+ players have survived the last season and a half without going on the DL.
Payrolls continue to grow, and so too with it does the cost (both on the diamond and on the balance sheet) of injury. Forward thinking franchises will move to reduce these losses any way they can, and will benefit accordingly for relatively little investment.
Grue said 07/10, 01:46 PM
Injuries are unpredictable, even with prevention and rehabilitation. Major league medical staffs are already some of the finest in the world because the ownership and players can afford the best private care. Nutrition and training staffs are extremely knowledgeable. The physical requirements of MLB players are understood, rigorous, and repetitive. If someone could gain a medical advantage short term, it will be minimal at best.
We are on the brink of the next "Moneyball". That advancement is predictive analytics through Simulation.
All of the pieces are in place.
The PITCHf/x system creates precise pitch data. For example, leading PITCHf/x Analyst Mike Fast was recently hired by the Baseball Operations department of the Houston Astros.
Advanced hitting and fielding results are tracked by sabermetric analysts.
Sophisticated video games with amazing realism entertain fans at home
Cloud data technology, high-end graphics+physics modeling, and real-time data processing are at our fingertips.
Now put it all together! Before making major decisions, run simulations. Play 1000s of games, situation tactics, multiple seasons virtually to evaluate risk.
MagicSpecs said 07/11, 07:33 AM
I am interested to hear how you believe the use of simulations could be translated into improved results (beyond the depth of player analysis already in use today), and how much of an edge this could actually give a team.
In my first argument I highlighted the frequency of injuries amongst the baseball elite. The cost of even a short DL stint is measured in millions of dollars, and this is why teams will actively move to reduce their losses.
There are already two high profile cases of injury mis-management this season. Matt Kemp and Evan Longoria have needlessly lost playing time as their injury rehabilitation programs both reaggrevated their original injuries.
Is this acceptable in today's game? No. It shows that the current systems do not work. Teams need to optimise injury treatment by investing further in their medical teams - investment that would be repaid in full by results.
Teams may also look to reduce incidental injury risk by managing player behaviour. Higher risk plays such as outfielders jumping into walls and batters hitting catchers at home plate could become restricted to close games and postseason play, reducing the chances of player injury.
Grue said 07/11, 09:45 AM
I'm glad you asked about the benefits of predictive simulations. We're very good at analyzing past behavior. We're very good at creating reasonable representations of prior seasons performance. I played with strat-o-matic lineups as a kid that did the same thing.
The military has used predictive models for decades for reducing risk in procurement decisions (buying billion $ systems), operations planning (saving lives), and more recently counter-terrorism, and homeland security.
Simulations have been tried in MLB before for decades. But those were the equivalent to Atari compared to today's X-Box.
Follow the bouncing Moneyball and look to the Society for American Baseball Research (SABR), MLB, and Bloomberg Sports. These groups hosted the first SABR Analytics Conference (http://sabr.org/analytics) in March 2012.
A competition was held to evaluate "whether Washington should buy or sell certain players such as starter Edwin Jackson or Cincinnati Reds second baseman Brandon Phillips, or stand pat and do nothing at the break." Speakers included agents, MLB presidents, and GMs.
MLB is starting to buy in. We're on the edge of an explosion and it's gonna be big.
MagicSpecs said 07/11, 12:32 PM
Whilst the thought of using projection software to determine personel strategy is intriguing, there are problems with suggesting that this is the next major advancement in MLB.
The human element of the decision making process will never be replaced, and projection data will only offer a compliment to existing statistical and scouting information.
Improving injury management could have significant impact on team performance. Reducing or avoiding just one injury per season could have a major impact on team results.
Software projections will never be perfect. In fact, it's questionable how accurate it could ever be - just the same as conventional scouting.
Injury management helps to maximise performance you already pay for and control. If you keep your best players fit and playing the results are undeniable.
Even if software projection software were available tomorrow it would be years (at best) before it could be trusted.
Injury management can be improved today. This very second Matt Kemp and Evan Longoria need not be on the DL. The time is now for Injuryball to provide an edge for teams willing to make the change and investment.
Grue said 07/11, 04:00 PM
Technology advancements will happen before medical advancements:
- the technology already exists in other industries
- The barrier is the same as Moneyball: old school culture of human element and eye-ball tests.
- The advantage is the same as Moneyball: Small-market teams are more willing to invest to compete with the large market salaries. Simulations cost less than high-$ salaries on under-performing players.
Predictive analytics are stat based, including medical risk. You can run scenarios that take key stars out of your lineup and project which farm-system or free-agent guys may replace them and the potential lineup shifts.
When running predicitve models, you're not looking for 100% accuracy. You look for relative not absolute impact: i.e. does your performance go up or down?
The human element of injuries can only be predicted with high uncertainty (unless you are Chase Utley). If some miracle method of injury management occurs, all teams will immediately adopt it and no advantage is gained across MLB.
MLB owners are already showing interest in advanced analysis. It is the next Moneyball.
|3||JimJ77 Can't make it up||130||53||26||68.4%|
|7||PCH- 2-1, Heat?||927||1537||406||39.4%|
- Fantasy baseball Weekly Planner: Upgrade your roster with Memorial Day sale
- Fantasy baseball Prospect Watch: Miller looks like an ace
- Fantasy baseball Stat Focus: Cabrera's wOBA proves dominance
- Fantasy baseball Trade Tips: It's time to sell Rickie Weeks
- Fantasy baseball Roundtable: Is Cardinals' pitching slowing?
- Awful Announcing
- Free Darko
- Pro Football Talk
- The Big Lead
- Joe Posnanski
- The Sporting Blog
- Big League Stew
- Bugs and Cranks
- Every day Should Be Saturday
- Mr. Irrelevant
- With Leather
- The Sports Hernia