Is StrokesGained Putting misleading?

[ZappyAd]

I started thinking about this after seeing a tour player being interviewed where he said he had a great round putting.  The analyst back in the studio then pulled out his SG Putting stats and said it was about average.  Made me wonder how the tour player who spends all his day looking at stats and practicing would not know what good looks like.  Ok, so sometimes the brain tricks us into thinking we are better than we are, but maybe the stats aren't telling the whole story.

My guess is that because SG Putting only takes distance to the hole into account it is missing a big part of the equation, namely 'magnitude and direction of break'.   This means that a player can think they putted great because the made a lot of big breaking putts (and they actually did) even if SG Putting puts them at average.  

As an example if the tour average from 8 feet is 50%, I think that this probably breaks down to something like (made up numbers for illustration)

- tour average for a dead straight putt from 8' = 75%;

- tour average for a stinky left to right 3' break from 8' = 25%.  

So the break makes a big difference to your chances of holing the putt (obviously).  It then also makes a difference to putting a rating on how good your approach shot was.  Hitting a wedge to 20' with a straight putt slightly up hill is much better than 20' downhill with a huge break.

Imagine 2 pros (A and B) playing together on the same course.  They both leave themselves exactly the same distance from the hole on every approach shot and they both make the same number of putts.  Player A always ends up with the straightest possible putt.  Player B always leaves a putt with as big a break as possible.  At the end of the round their strokes gained for approach and putting are going to be the same, but Player B will have had a much better putting day.  Player A will have had a bad putting day but a great approach day.  

On average relative to the field etc etc you get a mix of good leaves and bad leaves on your approaches so there is probably a tour average for the 'amount of break' the SG Putting number reflects.  But if you have an off wedge day where you leave the ball close but in the wrong spot all day but putt great to save your round, it doesn't look to me like Strokes Gained will reflect that.

This is obviously an extreme example but I think it shows that SG around the green isn't capturing enough information to definitively say whether a player has a good or bad day putting.  So maybe the tour pro does know something that the analyst doesn't.

[dennyjones]

Interesting thought.   

[David in FL]

Any stat can be misleading if context isn't considered.  The guy who had 32 putts and hit every green may have had a much better day putting than the guy who had 28 putts and missed every green. 

The strokes gained stats do a better job capturing the overall picture than others, but in the end, there's always going to be outliers and circumstances that aren't captured in the number.

[saevel25]

It could be that pro's misread putts more often than you think. That doesn't make the putt is harder, it just means they need a better way of reading putts. 

 

Edited July 1, 2017 by saevel25

[RandallT]

Yah I think it's an interesting topic. I believe I have read somewhere (LSW?) that distance of putt is far and away the most significant factor.

Intuitively we might think the amount of break is significant for them, but I would like to see the "stats" on 8ft side hill putts with 3% slope or whatever vs straight.  My guess is that the difference is marginal- could be wrong. 

The other thing at play here is that each week, the PGA does an adjustment to the SG. They don't go straight from Mark Broadie's tables, where 8ft =50%.  They do an adjustment for how the field performed that week. 

As I recall on the PGA site, they show the fancy adjustment formula but it boiled down to using baseline data to get the number and adjusting for that week's performance of the field.

i mention this because it could be that the greens played easy that week (relatively) so a player felt like he was on fire. Unfortunately, other players played well too and the adjustment kicked in. 

Hence the player who thought he was hitting long putts was just doing what everyone else was doing too.  Possibly another explanation for the interview you saw, so I throw it out there. 

[iacas]

12 minutes ago, RandallT said:

Intuitively we might think the amount of break is significant for them, but I would like to see the "stats" on 8ft side hill putts with 3% slope or whatever vs straight.  My guess is that the difference is marginal- could be wrong.

The difference is not too large, correct. It's not insignificant to have a sidehill putt versus a fairly straight uphill putt, but it's not 20%, either.

12 minutes ago, RandallT said:

The other thing at play here is that each week, the PGA does an adjustment to the SG. They don't go straight from Mark Broadie's tables, where 8ft =50%.  They do an adjustment for how the field performed that week. 

Right. Some greens are trickier than others. They take the field average. Sometimes that 50% number is 9', sometimes it's 7' (with sliding scales for everything else).

---

To the OP, yes, if you have a small sample size, it can skew things. That's always been true in statistics, and 28 putts or whatever is a small sample size. You could find yourself with a bunch of big breaking putts, relatively speaking.

Remember, too, that 28 isn't even really the number. Rule out all tap-ins to three-footers as being irrelevant and you're down to fewer than 20 putts that really count.