CBB '24 - Week 5 Recap: The Anatomy Of A Bad Line
Examining how we got Houston-Kansas, Colorado-Utah and Tennessee-Kentucky wrong; HQ retreats in shame to ML-only markets for Week 6; Leaderboard, Responses Points Percentages inbound...
Well done. You won so many points off HQ this week that we're dragging our tails behind us and pondering a Week 6 card that has only "to win" markets. Congrats are deserved by continuing leader Michael "Seymour" Hirshenson, and his trailing coterie of proximate followers.
To clarify for those unaware, we are forced to put out lines before the market makes them, in order to give you all enough time to fill out your weekly cards.
If we waited to send the card to you until we could copy public lines, you all would have from about 1pm PT Friday until 9am PT the next day to fill out your cards.
It is already extremely hard to get a critical mass of people to fill out a detailed form with points allocations in any span of time, let alone during 12-ish waking hours over a weekend.
As such, we're guessing as to what these lines will be when we send out the card, and sending them out dangerously early in order to give you, the good people, an extra 24 hours or so of time to consider the games at hand.
What are some pitfalls that come with that? Well, some misses, for one.
Let's take three discrete examples from this weekend and get some key learnings (i.e. self-roast).
1) Analytics Sites Over-Value Some Teams
We opened the weekend by nut-tapping ourselves: We listed Kansas as a 4.5-point home dog against Houston in the contest. Of course, that line stayed there, because our tournament does not feature dynamic lines.
This opening guess was based off of several different analytics sites that projected Kansas as anywhere from a 3 to a 5 point underdog. This was very similar to the projection Kansas had as road underdog at Hilton Coliseum in Week 4, a projection that was bang on the money.
But when Houston goes flat, they go real flat (see: 2023 NCAA Tournament vs. Miami). And we would have done well to remind ourselves that Houston is the darling of the analytics websites we use for projections. Thus, maybe we should've shaded down a bit.
The result was you all feasting like wolves on a line that was inflated to begin with, and then due to real-world movement, three points off of the closing line.
96%, or 9,650, of the points allocated on this game were allocated on Kansas to win.
After our line came out, Kansas opened in the real-world as a 3-point home dog and was bet down to a 1.5-point home dog, so we were 3 points off.
Kansas demolished Houston with a complete performance from start to finish.
Rock Chalk, and all that. Our mistake is your gain.
Ironically, we wrote in last week's email that we were looking forward to everyone picking Kansas and Kansas winning. So, I guess we did anticipate how the market would react but just put out a bad line anyway.
2) Last-Minute Injury News Can Bite You
Colorado @ Utah was a miniature version of the same Houston-Kansas mis-calculation, with personnel issues thrown in.
For reasons that still boggle us to this day, this is a Utah crowd. It's the worst-kept secret of Robn. So, we had to put this blockbuster game on the card.
But how do you put out a line for a game when you're not sure if a team is going to have its star guard or star big man play?
Separately, Utah in its previous two games was also projected as a small favorite, and it lost both by a combined 43 points, hurting deeply from the absence of Rollie Worster and Lawson Lovering. So, how do you account for future performance of a team with recent performances that are in the extreme outlier category relative to projections?
Robn's answer to both questions was, "who knows!?" and we shot out a downward-shaded line that accounted for at least one Ute player being out and the team's recent vulnerability.
Analytics sites projected Utah as anywhere from a 1 to 3-point home favorite against Colorado. We put them as the equivalent of a 1.5-point underdog.
But after we did this, news crept out that Lovering was going to play (and he does, helping anchor the Utes interior presence).
Publicly, Utah opened right in line with the analytics projections: -2.5. It was then bet up to -3.5. Meanwhile, Robn was sitting there with egg on its face, treating Utah like a sight underdog rather than a slight favorite to try to account for the above factors, at least one of which is no longer a factor.
You all saw this and ...94% of points allocated on this game were put on the Utes.
Utah almost blew it anyway, letting a late lead evaporate, but would go on to win by 5.
3) People Love Picking Name-Brand Schools At Home, ESPECIALLY Underdogs
Tennessee at Kentucky was arguably the most meaningful game of the weekend and the two teams ended up entertaining the good people by playing a 2013 Big 12 football game.
Analytics sites projected the game anywhere between Tennessee as a 1- and 5-point favorite. The Vols are red-hot and while on the road in a hostile environment, they are an older and more well-rounded team (in HQ's opinion) than Kentucky. Robn landed right in the middle of these numbers, making the line the equivalent of Tennessee -3.
This one, we didn't actually get too bad. Tennessee opened -2. What we didn't get right was predicting where the line would end up in the public markets. Over 70% of real-world money and tickets consistently hammered the Wildcats on Friday afternoon and Saturday morning, and the line flipped to Kentucky as a 1.5 point favorite.
Robn's contestants mimicked the real world in this regard. 76% of all points allocated on this game were allocated to Kentucky.
Why?
People can't resist blue-blood teams at home. It is built into the simp DNA. But it's Kentucky! Kentucky's always good, right? A blue-blood team playing at home and as an underdog? That grade of catnip can be served a la carte.
Tennessee would end up winning by 11 points, and our game cap assessment was correct. But we're trying to mimic the public markets; we should have anticipated this and opened at a Pick'em, at least.
The total? That we actually got reasonably well. Public markets opened at 157, and we listed 155.5.
The combined 195 points was just reflective of defense-optional hype in Lexington.
Please note that Bobby Dillingham came off the bench and SHATTERED his EXTRA market (to score at least 17 points - 1.2x) by putting in 35.
Tragically, none of you picked it...
Week 5 Results and Leaderboard
Leaderboard Headed Into Week 6
Click here for the leaderboard
Click here to see all responses
Click here to see all graded responses
Allocation Of Bonus Points (see Columns AV and AW in the Null WK 5 Tab)
Click here to see the allocation of all Bonus Winnings
