Strokes gained for simulator golf, explained
Strokes gained measures a single shot against a benchmark: the average number of strokes it takes a reference group of golfers to finish the hole from where your shot started, compared with the average from where it finished. If the shot moved your ball to a spot worth more than one stroke of progress you gained strokes, and if it moved you to a spot worth less than that you lost them.
The measure was developed by Mark Broadie at Columbia University, who applied it to the PGA Tour's ShotLink shot-location data and published the analysis in the journal Interfaces. Broadie's earlier work called the same idea shot value. It is now the standard way tour performance is broken apart, and there is nothing in the arithmetic that requires you to be playing outdoors.
The baseline is the entire idea
Everything rests on a benchmark function that Broadie writes as J(d, c), where d is the distance still to the hole from where the ball currently sits and c is the condition of that lie: green, tee, fairway, rough, sand or recovery. J is the average number of strokes a golfer in the reference group takes to hole out from that position. Broadie estimated it from a database of more than eight million PGA Tour shots hit between 2003 and 2010.
Some real values from that estimate. From 100 yards in the fairway, a tour player averages 2.80 strokes to finish the hole. From 100 yards in the rough it is 3.02, and from 100 yards in sand it is 3.23. That gap is the honest price of the miss, expressed in the only unit that matters. Off the tee on a 400-yard hole the average is 3.99. On the green, the tour averages 1.05 putts from three feet, 1.50 from eight feet, 1.61 from ten feet, 1.87 from twenty feet, 1.98 from thirty feet, and 2.21 from sixty feet.
Notice that the putting numbers are not whole strokes. That is the point. A tour player holes an eight-footer often enough that the average sits at 1.50, so making one is worth half a stroke and missing one costs half a stroke.
How a shot gets a number
The formula is a subtraction. Take the baseline where the shot started, subtract the baseline where it finished, then subtract one more for the stroke you used:
strokes gained = (baseline at start) - (baseline at finish) - 1
In the example Broadie gives, a long par 3 takes the field an average of 3.2 strokes from the tee. A player hits the green to sixteen feet, where the baseline is 1.8 putts. That tee shot is worth 3.2 - 1.8 - 1 = +0.4. He then leaves the birdie putt an inch short, finishing at a tap-in worth 1.0, so the putt scores 1.8 - 1.0 - 1 = -0.2. Missing a sixteen-footer is not a full lost shot, because the field only holes about twenty percent of them. The tap-in itself scores 1.0 - 0 - 1 = 0, which is correct, because sinking a one-inch putt neither gains nor loses anything.
A worked example with real data
Angel Cabrera played the Deutsche Bank Championship at TPC Boston on 4 September 2010 and took 26 putts. Against the tour average of 29 putts per round, that reads like an excellent day on the greens.
It was not. His first putts started an average of 9.6 feet from the hole, far closer than the tour average first-putt distance of 17 feet. Given the actual distances he faced, the tour baseline for that set of putts was 23.2 putts. He took 26. His strokes gained putting was 23.2 - 26 = -2.8, a loss of nearly three shots to the field.
The raw count flattered him for two reasons. He had two greens where he never putted at all, which lowers a putt count without saying anything about putting. And his approach play left him short putts that the field usually makes. He then three-putted from 22 feet and again from 20 feet, and missed from five feet and seven feet.
The mirror case ran the day before at the same event. Ian Poulter took 32 putts, three more than the tour average, and gained 2.3 strokes. His first putts averaged 32 feet. On one hole he two-putted from 90 feet, where the tour average is 2.4 putts, and banked 0.4 strokes for doing something entirely unremarkable.
Why it carries indoors
The calculation asks for three inputs and no others: where each shot started as a distance and a lie, where it finished, and a baseline. Nothing in that list is a property of being outside. A simulator round produces exactly the start-and-finish pairs the formula consumes, which is why the concept transfers without modification.
The baseline also does not have to be the PGA Tour, and for most of us it should not be. Broadie fitted a tee-shot baseline for golfers who average 90 for eighteen holes and got an average score of 2.79 + 0.0066d, against 2.38 + 0.0041d for tour professionals. Read the slopes: every extra 100 yards of hole length costs a 90-golfer 0.66 strokes but a tour player only 0.41. Measured against a reference group near your own standard, strokes gained describes you. Measured against the tour, it mostly tells you that you are not on the tour.
What it will not tell you
It is a measure of outcome, not of cause. Broadie is explicit that strokes gained scores a shot by its starting and finishing locations, so the swing that produced the result sits outside the measure entirely. A shot worth +0.5 is not evidence of a good strike; it is evidence of a good resting place.
It is also relative. Change the reference group and every number in your round moves. And it is sensitive to conditions, which is why the tour adjusts each player's strokes gained for a round by the field average for that round, and why Broadie's own golfer rankings adjust for course and round difficulty before comparing anyone. Those rankings were built from players with at least 30 rounds in a season, which is a useful hint about how much play it takes before the numbers settle into something you can trust.
One round is a data point. A season is a diagnosis.
GSCaddie computes strokes gained for your simulator rounds.