Kyle Lowry - 9.1.15

There was a 'trend' highlighted on Twitter yesterday, which was that, and I kid you not:

Since Jan 9, 2015, in games where the Toronto Raptors are favourite, coming off a loss when they were favourite, and when more than half of Kyle Lowry's made field goals were 3 pointers.

The 'trend' was that of the 17 games since that memorable January day of 2015 (no, me neither), the Unders had a 16-0-1 record.

As often happens with these 'trends', the act of publishing it is all that is needed to doom it, and the game last night (Raptors @ Charlotte Hornets) went comfortably Over.

The thing is, such 'trends' are merely curiosities, the result of some presumably time-consuming data mining. 

Maybe these things are interesting to some, but as indicators of where a market offers value, they are dangerously useless.

Of course nothing happened in Kyle Lowry's life on January 9, 2015 whereby the relationship between his two and three point baskets suddenly took on a mystical importance.

Look at enough combinations of data and of course you will find examples like these, but if there is no rationale behind them, they are worthless. Not to mention that checking a database daily for five years and having just 17 selections probably isn't the best use of your time. 

No one on the planet is looking to back Overs "in games where the Toronto Raptors are favourite, coming off a loss when they were favourite, and when more than half of Kyle Lowry's made field goals were 3 pointers" and thus pushing the value to Unders.

There's another one tonight if anyone is interested:

Since December 2019, a road team that is off a win in which they were not winning at the end of the third quarter and they allowed fewer than 28 points in the fourth. It is 20-0 ATS and it is active tonight!

This one is actually slightly interesting because it's based on a solid idea - that there might be some carryover when a team finishes strongly and wins. 

However, I'd expect the value to be on the side of opposing these teams, since recency is such a strong bias. I'm not interested in the number of points that the opposition scored in that final quarter, nor do I believe that the night of November 30th was momentous. That's just fluff to make the 'trend' look more glamorous. 

So as I do, I took a look at teams who came from trailing after three quarters to winning the game and here's what I found going back to the 1995 season in regular and playoff games:

Over the last one and a half seasons, backing road favourites in this position (regular season) has a 31-21 (59.6%) record.