You probably saw the statistic that 77% of Game Three winners in series that were tied 1-1 have gone on to win a seven game playoff series. It was discussed in this article on ESPN previewing last night’s Heat-Pacers game and has also been mentioned on telecasts.

Let’s put that in perspective, though. What does that really mean, and how does it relate to the Heat-Pacers series, where the heavy favorite entering the series is now down 2-1 after a road loss in Game 3?

First, a quick thought experiment. We know that, if a series is tied after two games, it must go at least five games. If we were randomly shown the results of only one game between games 3 through 5, what are the chances we would pick the series winner? The answer, assuming the two teams are roughly equal, is approximately 69% of the time. I got this by working out the likelihood of winning the series for a team with a 2-1 lead, assuming they had a 50% chance of winning every game thereafter, and seeing who ultimately got to four wins first (with home court advantage, two equal teams might alternate being favored, but both should have the same number of home games if it goes to a game 7).

So, assuming that 69% is our theoretical result just knowing who won one of the next three games at random, that 77% doesn’t sound quite as good. Still, I went back through the NBA archives and it turns out that the Game 3 winners have been slightly more successful than knowing Game 4 or Game 5 winners.

Overall, Game 3 winners in a split series after the first two games won 76.5% of the series. Game 4 winners won 65.4% of series, and Game 5 winners won 68.6% of series. Combining all three games, randomly pull one of those three out of a hat after knowing the series was tied, and you would have predicted the winner 70.2% of the time. Very close to my theoretical 69% assuming the two teams were equal and you knew the winner had to 2 wins out of 3 known results.

Of course, that result may not be wholly applicable to the Heat. In this case, Miami is the favorite, and at least based on the regular season, this would not be a 50/50 series. Of course, Chris Bosh’s injury may change that dynamic.

I went back to the 1990 Playoffs and found all seven games series where  a) the teams were tied after two games, b) the home team (worse seed) won game 3, and c) the better seed was at least 2 points better in simple rating system based on regular season point differential results. There were 16 such series prior to this one that met that criteria, most recently of course being the Bulls against the Sixers in the first round this year.

The “better team” came back and won the series 8 times. Three of those times, the better team won the series in six games, by sweeping the next three. The other five times, they forced a game 7 and won at home.

The “underdog” held on to win the series 8 times. All eight series wins were in six games, which highly suggests that the Pacers need to close it out before the Heat get back to a game 7.

Does that seem about right? I think it does. Knowing that a heavy favorite is now down 2-1 in a series gives them about a 50/50 shot at coming back. Of course, the injury factor has crept up in some of those series as well, including Derrick Rose this year. The Bosh injury might actually make the Heat a slight underdog for the overall series. However, Miami definitely has a better chance than that 77% number for the Pacers winning game 3 would suggest.

[photo via US Presswire]