Kirk's Hammer: When CLV Isn’t the Right Tool

Anyone who has read my newsletter knows I’m a closing line value maxi. I evaluate nearly all my bets using the closing line, and rarely do I think a bet that doesn’t close well was a good one.

I’ve covered the idea before — like in my politics article, where I wrote about how to evaluate bets when CLV isn’t perfectly defined. But today I want to go deeper on situations where the close is actually a bad metric to use when evaluating your bets.

There are two spots that stand out: lookahead lines and injury bets.
 

• With lookaheads, there’s simply too much new information between when you bet and when it closes to fairly evaluate using CLV.
• With injuries, you’re basically betting on the percentage chance a player is in or out. If you thought there was a 50% chance the best player in the league (worth six points) played, and you got one point of CLV when he did, that doesn’t make it a good bet. 
   

This week on Twitter I posted two polls (and Peanutbettor posted another) on exactly this topic. I want to go through them and explain why I think the bets were good or bad.

                            ________________________________________


Example 1: Liberty -6 with Sabrina Questionable

Tweet: Bet NYL -6 today, Sabrina questionable and thought she had a legitimate chance to play. Play had tons of resistance all day, was around -6.5, she gets ruled out, closes -6. This bet was: A. Good B. Bad C. See results.

Poll results:

-32.7% good
-36.4% bad
-30.9% wanted to see results

First off — if this was just some degenerate clicking random buttons, the bet is bad. But assuming we’re talking about a capable bettor here, there’s more to it.

The most important thing to me in injury situations is what the number was before the announcement. That’s like the “close before the close.” Everyone knew Sabrina was questionable, and the market landed on Liberty -6.5 while baking that in.

That’s the best argument against the bet: directly prior to the injury announcement, -6 was basically neutral EV. On the other hand, some big market players just don’t touch these spots until there’s certainty — not because they think the line is wrong, but because they know their strength is in making numbers, not predicting injuries.

Now, the argument for the bet: The game closed -6, which means no Sabrina made Liberty -6 roughly -4.5% EV. But I have Sabrina worth about three points to the spread. If she played, that bet jumps to +16% EV. She only had to play around 25% of the time to make -6 a good bet.

Given WNBA “questionable” tags usually play more than 50% of the time, and even though Sabrina’s situation looked worse than a coin flip, it was still clearly north of 25%. That makes the bet good.


                            ________________________________________


Example 2: Sparks +7.5 with Allisha Gray Questionable

Tweet: Bet LAS +7.5 today, thought there was a chance Allisha Gray was out and I made the game 7 with her in. Similar to yesterday line moved very small my way and was painted 7 most of the day. Now Gray is out and line will close ~5. This bet was: A. Good B. Bad C. Neutral D. See results.

Poll results:

-74.6% good
-10% bad
-15% either neutral or see results

This one looked great to most people. But it was almost the exact same situation as the Liberty bet — and maybe even worse.

The line ticked half a point before news, same as the Sabrina spot. I had Gray’s status priced about the same probability (40%). When she was ruled out, the line moved two points in my direction. But if she had played, it probably would’ve gone to 7.5 or 8.

By the math, this was likely worse than the Liberty bet — even though the injury news went my way, so the close looked much better.

                            ________________________________________


Example 3: Florida +7 vs LSU Lookahead

Tweet from PeanutBettor: YTD Florida in bottom 5th percentile of expected performance. LSU top 10th percentile. Let’s say YTD performance worth approx six points. Preseason you bet Florida +7 vs LSU. Now, line is 9.5. Good bet?

Poll results:

-26.9% yes
-42.4% no
-30.7% see results

This one is totally different. Peanut bet Florida +7 preseason. A few games in, Florida had looked awful, LSU had looked great, and the line drifted to 9.5.

CLV is not very useful here. Too much info came in between when the bet was placed and when the game is played. But that doesn’t make the evaluation easy. Lookahead lines can be argued in different ways.

When I bet NFL week-ahead lines, I’m trying to capture what the number will be if both teams perform exactly to expectation this week. If I bet Chargers -1.5 vs Denver in Week 3, and then the Chargers lose by 70 in Week 2, the line might flip to Denver -3. That doesn’t make my bet inherently bad — because I wasn’t trying to predict the Chargers’ Week 2 result. If I wanted to do that, I’d have just bet Week 2.

But Peanut’s case is a little different. This wasn’t a one-week lookahead; this was a college football line weeks into the season. In that context, you might actually be betting on performance outcomes too — e.g., thinking Florida would surprise or LSU would disappoint.

Even then, I’d still argue the bet was good given the facts he laid out. Florida being bottom-5% and LSU top-10% was probably unprojectable. And if that gap was truly worth six points and the line only moved 2.5, Peanut probably got in good — even if the ticket looks bad on paper.

                            ________________________________________


Closing Thoughts

In these spots, you need nuance to evaluate your bets. Too often I see people get a close in their favor and assume the bet was good, but when it goes the other way they hand-wave it — “oh, injury news went against me” or “I couldn’t have possibly known Florida would be this bad to start the year.”

If you’re going to use that kind of cope for the bad bets, you need to apply the same lens to the ones where you ran good. The best way to evaluate these is to track them all — along with the assumptions you made — and then look back at how well you’ve actually done with these specific bet types.

That’s where the truth comes out.

                            ________________________________________
 

Check out Kalshi and get signed up today: https://kalshi.onelink.me/1r91/hammer