Analytics

[RiverDog]

It's a method that football is starting to embrace, so I thought it worth a discussion.

Analytics is nothing new to sports. Although never called by its current name, baseball has been tinkering with it for decades ever since the St. Louis Cardinals employed a shift against a slugging left handed hitter named Ted Williams, the American League MVP, in the 1946 World Series. They stacked the left side of the infield and outfield against the dead pull hitter, the best hitter in baseball. Williams hit nearly 150 points below his season average and the Cardinals beat the heavily favored Red Sox and won the series.

Baseball is much more suited to analytics than football. Every hitter has a batting average, every pitcher an ERA. Relative to any individual statistic kept in football, such as a completion percentage or yards per carry, baseball's statistics are more objective and less likely to be influenced by other factors.

My personal take is that it's currently getting used in inappropriate situations. For example, on a 4th and 2 situation in the first quarter and at the opponent's 25 yard line, analytics would determine that a team try for the first down vs. kicking a field goal. The problem is that it does not incorporate subjective factors, like the strength of your running game, the defense they're running against, wind and weather conditions, and so on. Those types of factors are usually not in play in a MLB game. There's no scatter diagram showing a hitter's tendencies. It's all based on an accumulation of multiple team's success rate. My opinion is that analytics are a useful tool for a coach contemplating a decision, but they should not replace the good ole human intuition of an experienced coach that knows his team's and his opponent's strengths and weaknesses

Comments?

[NorthHawk]

Analytics are statistics, so they can lead someone astray if they are used as the only factor. They are good to show tendencies like you said,
but Football is such a game of emotion and momentum as well as matchups that the statistics can in some cases be only minor factors in how
any play could turn out. As well, if you know the other teams stats vs plays and they know their own tendencies, they can counter it by breaking
their tendencies and stopping what you are trying to take advantage of. It's still a giant chess game, but I'm not so sure analytics should be
anything more than an aid or part of the decision making process. It shouldn't dominate those decisions.

[RiverDog]

Analytics are statistics, so they can lead someone astray if they are used as the only factor. They are good to show tendencies like you said,
but Football is such a game of emotion and momentum as well as matchups that the statistics can in some cases be only minor factors in how
any play could turn out. As well, if you know the other teams stats vs plays and they know their own tendencies, they can counter it by breaking
their tendencies and stopping what you are trying to take advantage of. It's still a giant chess game, but I'm not so sure analytics should be
anything more than an aid or part of the decision making process. It shouldn't dominate those decisions.

I'm in complete agreement. But it seems to me like they're being used more and more to make game decisions, in particular, more teams are going for it on 4th and short rather than punting or kicking a field goal based on analytics. I'm a little more old fashioned and feel that you always take the points in the first half of a game regardless of what analytics say.

Plus it brings up another subject, something that was broached by the broadcast crew in last night's SNF game. If it's 3rd and 2 at your opponents 22 yard line and analytics says that on 4th and 2, you should go for it, how does the knowledge that you would go for it on 4th and 2 influence your play call on 3rd and 2? Do you want to take a shot at the end zone?

There are some situations in football where analytics can play a major role. Point after touchdown attempts comes to mind. If the pre-PAT score is 14-7 in the 2nd half, would analytics say that you should go for 2 and make it a two score game where the failure to convert would still leave you with a one touchdown/kicking PAT lead?

[NorthHawk]

I think that's where the human element comes into play. How well your team is playing, not only during the game but on that particular drive will have a larger
role in the decision process than the analytics. Because it's just one of the tools for a coach to use, most would probably take into consideration how well their
team is doing at that moment - maybe over what the numbers say.

A lot of this comes from the college game but not all of it is transferable to the NFL where there is a much more consistent level of talent throughout the league.
Therefore, there will be some adoption of these tactics, and it may take hold in a larger way for a few years but I believe it will fade to being just some more
information to consider. Of course, some coaches are more prone to higher risk strategies than others, but unless they are consistently successful using these
tactics, I'm not so sure it will be a dominant factor over the long run.

[RiverDog]

I think that's where the human element comes into play. How well your team is playing, not only during the game but on that particular drive will have a larger
role in the decision process than the analytics. Because it's just one of the tools for a coach to use, most would probably take into consideration how well their
team is doing at that moment - maybe over what the numbers say.

A lot of this comes from the college game but not all of it is transferable to the NFL where there is a much more consistent level of talent throughout the league.
Therefore, there will be some adoption of these tactics, and it may take hold in a larger way for a few years but I believe it will fade to being just some more
information to consider. Of course, some coaches are more prone to higher risk strategies than others, but unless they are consistently successful using these
tactics, I'm not so sure it will be a dominant factor over the long run.

I would think that if talent was more evenly distributed in the NFL than it is college that it would make analytics more applicable, not less. Results would be more predictable.

Wasn't it Chip Kelly that always seemed to be going for it on 4th and short from his own 30 or 40 yard line?

[trents]

I'm coming from a different angle, not in-game decision making by the coaching staff but from the announcers and color people in the booths. I grow weary of the constant use of analytics by announcers and color analysts. So much of what they spout is so very trite that makes some kind of point that no one cares about. Aaron Goldschmidt of the Mariners drives me nuts with that. They are constantly proclaiming stuff like "this is a new record for left-handed quarterbacks who have played less than three years and who have green eyes." It's like they can make a record out of anything. Who cares! Maybe that's not exactly analytics but the incredible amount of data from digital record keeping makes for a lot of trite statistical commentary.