What cricket can learn from baseball's review system
The Decision Review System was adopted in cricket in 2008. The system currently comprises two parts - a set of technologies, and a set of protocols governing how the technology is to be used by umpires and players. It allows for two types of review - a review by players of an umpire's decision on the field, and a review by umpires on the field (limited to specific types of decision) before they reach a decision. Major League Baseball adopted a limited review system in 2008 as well. For the 2014 season, this was expanded substantially.
It is useful but not essential to understand all the intricate rules of baseball in order to understand how their review system compares to the DRS. Each major league game is managed by an umpiring team (known as a crew) of four to six umpires. Most regular season games have four umpires - one on home plate, and one each on first, second and third base. Two extra umpires are typically used in the post-season games and are positioned in the outfield. The most experienced umpire in each crew is known as the "crew chief". The umpire at home plate for each game is known as the "umpire-in-chief".
The role of the manager of a team in baseball is far more important during a game than that of a head coach (which is the equivalent position in cricket). The manager wears the same uniform as the players and sometimes calls each play from the dugout. The normal practice in baseball is for the catcher (the baseball equivalent of a wicketkeeper) to "call" a pitch. This would be like MS Dhoni telling R Ashwin what variation to deliver at what pace in what line, every ball. Catchers often look for instructions from the manager. For the batting side, batters who have reached first or second base are often given instructions on a pitch-by-pitch basis as to what they should try to do (for example, whether or not they should attempt to "steal" a base). All this communication occurs through elaborate, secret signs.
The crew chief and the two managers lie at the heart of the new review system in baseball. When the review system was expanded in 2014, it allowed teams to review certain types of decisions. A decision may be reviewed by a manager from the dugout.
If you have played cricket where the stump at the non-striking end was a single stone placed on the ground, and run-outs could be effected by catching the ball and touching the runner with your hand (while holding the ball) before they crossed the crease, you were basically using a "tag-out" from baseball.
Suppose the umpire at second base rules a runner out to a tag at the plate by the second baseman. The batting team feels that the umpire has made a mistake. The manager of the batting team walks out onto the field from the dugout, while an assistant calls the club house on the phone to find out if the umpire has clearly made a mistake: the replays are shown on TV while all this is going on, helping the management of the batting team decide if a review might lead to a reversal. Armed with this information, the manager, who has effectively been playing for time, requests a review or walks back to his dugout (in effect accepting the original call).
A decision may also be reviewed by the crew chief at his sole discretion. Each team is allowed one review per game. Teams are allowed one additional challenge should they use their challenge successfully. Should a team have exhausted this review, from the seventh innings onwards (a game has nine innings) they can request the crew chief (but not insist) to review a call. Typically crew chiefs tend to grant such reviews.
As with everything else in baseball, the performance of the review system has been measured meticulously by the official authorities and by serious fans. As I write this, 480 out of 1023 calls have been overturned, a rate of about 47%. A review is requested at the rate of one every two games. If you consider the times when the video analyst of a team recommends against requesting a review, then, according to Joe Torre, the reversal rate would be about 22%. Effectively the ability of a team to review video before requesting a review doubles the success rate of reviews. The umpires informally allow managers about 30 seconds to get a recommendation from their video analyst.
Many of these reviews are marginal calls. Three out of four requested reviews relate to force outs or tag outs (baseball's versions of the run-out). The video evidence has to conclusively show the original call was wrong. Calls of balls and strikes by the home plate umpire are not reviewable.
One interesting aspect of baseball's review system is that the baseball equivalent of the third umpire is situated in a centralised facility in New York City for all baseball games. This makes a lot of sense as over a dozen games are played every day during the baseball season (2430 games in the full regular season). This probably ensures greater consistency in decision-making by the reviewers of the video replay.
According to early reports, the instant replay has not improved behaviour in baseball. Managers have been ejected from games at about the same rate as before. Arguments with umpires (accompanied by temper tantrums) are a ritual in baseball and are often cheered or jeered by the spectators, depending on whether the home team's manager is involved or not.
The average review in baseball uses about 110 seconds. When the system was devised, it was estimated that this would be about 90 seconds. However, these are early days. It is far from clear whether this time spent is in fact spent poorly.
How does the review system in baseball differ from the one in cricket? What do the procedural differences amount to?
First, baseball allows teams to review decisions themselves before asking for reviews by officials. I do not suggest that the review success rate in cricket will improve by a factor of two (as it does in baseball) if the coach and video analyst are allowed to request reviews from the dressing room. But I do suggest that some of the more desperate reviews could be eliminated using this method. Limiting the number of reviews to one would also be more reasonable if teams were allowed to look at video before requesting a review. For this, cricket would have to exert tighter control on the broadcast than it does now.
Second, baseball allows reasonable discretion to the umpires and managers on the field about requesting reviews. Umpires are not undermined as they are in cricket. Team managers do not consult with the players involved before requesting a review. They rely on video.
The third difference, which may be the most significant of all, is that reviews in baseball do not involve verifying one conjecture with another. Only reasonably verifiable decisions are reviewed. The home plate umpire is still the sole judge of whether or not a pitch is in the strike zone. Strike zones are defined relative to batters. (See this wonderful story about Eddie Gaedel, the shortest player in major league baseball history.)
A team may, however, review if a batter is hit by a pitch. The cricketing equivalent of this would be for a team to be able to review whether or not there was an inside edge, or whether or not the impact was within the width of the stumps in an lbw decision, but not whether or not the ball would have gone on to hit the stumps had the pad not been in the way. With today's computer-vision technology, it is possible to model each player's strike zone on the fly, but professional baseball has refrained so far.
All sports have their own self-image. Cricket's self-image celebrates the autonomy of the individual player - especially that of the batsman. It is entirely consistent with this self-image that batsmen be allowed to request reviews without consulting anybody when they are given out. Baseball, on the other hand, lionises the manager, who picks the team, decides the batting order, decides when to change the pitcher, when to change a position player, what play to call during a game. The best player in a baseball team bats as often as any other player, unlike in cricket where the best player is slotted in at 3 or 4 and can play until the team's innings ends.
It is perhaps too much to expect a sport to change its basic character. But there are many lessons for cricket in baseball's adoption of instant replay. Perhaps the most important lesson is that if the ambition of the system is to end up with the correct decision, then using technology to help experts is a better approach than using it to second-guess them.