2025 NHL DRAFT

[Porter Stoutheart]

I think just to add to that, I've seen spreadsheets of draft success rates by position, I've even reproduced them here in the past, although I haven't gone looking for the source I used to quote in this post.

But the bottom line is you are going from about 30% to about 20% odds of hitting in the last 10 spots of the 1st round and early 2nd round. So you are talking about a 10% difference in odds. Think about that statistically...

1) say you pick #25, #30, #35...
Odds = 27% + 23% + 20% of getting a player = 70%

2) say you package #25+#35 to move up to #20...
Odds = 33% + 23% = 56%

There are lots of papers and studies of draft pick valuation these days, but I don't think there is any statistical argument which says trading up makes sense. It never makes sense statistically. All you can do is think you are smarter than everybody else and will beat the draft odds by getting a 17-18 year old player who in that part of the draft is probably 4-5 years from being able to play in the NHL, and thinking you are going to get the right guy, develop him right, he's not going to get hurt, all that. Which based on my experience is usually hubris. All you are going to succeed in doing is giving yourself a short-term dopamine hit for "getting your guy". In the long run, however, it is not a sound strategy.

 

[AtlantaWhaler]

I think just to add to that, I've seen spreadsheets of draft success rates by position, I've even reproduced them here in the past, although I haven't gone looking for the source I used to quote in this post.

But the bottom line is you are going from about 30% to about 20% odds of hitting in the last 10 spots of the 1st round and early 2nd round. So you are talking about a 10% difference in odds. Think about that statistically...

1) say you pick #25, #30, #35...
Odds = 27% + 23% + 20% of getting a player = 70%

2) say you package #25+#35 to move up to #20...
Odds = 33% + 23% = 56%

There are lots of papers and studies of draft pick valuation these days, but I don't think there is any statistical argument which says trading up makes sense. It never makes sense statistically. All you can do is think you are smarter than everybody else and will beat the draft odds by getting a 17-18 year old player who in that part of the draft is probably 4-5 years from being able to play in the NHL, and thinking you are going to get the right guy, develop him right, he's not going to get hurt, all that. Which based on my experience is usually hubris. All you are going to succeed in doing is giving yourself a short-term dopamine hit for "getting your guy". In the long run, however, it is not a sound strategy.

I'm terrible with statistics, but that's not how it works. You don't just add the percent. To illustrate, if you added a few more picks, it would be over 100%. I have no idea how it's done, but when looking up the first scenario, it came to 55%.

 

[Bringer of Jollity]

The higher theoretical upside of the player being traded up for often isn't worth the risk of reducing your two kicks at the can down to just one. If we're talking about trading up to top 10 or something that's a bit different, but how often do teams trade out of those spots?

 

[Porter Stoutheart]

I'm terrible with statistics, but that's not how it works. You don't just add the percent. To illustrate, if you added a few more picks, it would be over 100%. I have no idea how it's done, but when looking up the first scenario, it came to 55%.

They are completely independent outcomes. If you were asking what the probability of one or the other happening (but not both) in dependent cases you could do P(A OR B) = P(A) + P(B) - P(A & B) which would lower the sum. But that's not what we're asking here. There is no overlap whatsoever in these events?

If you have 5 late 1st picks... I see where you are coming from. Obviously there is a chance you f**k it up and get nobody.

Maybe .27 + .23 + .20 - (.27*.23*.20) is right? But then you are still at 68.758% instead of 70%. So it's not much different. It keeps you from going over 100%. But it doesn't change the argument. (In fact, it kind of strengthens it, since if you apply that to the 2nd case, you are at .33 + .23 - (.33*.23) = 48.41%, so you are just making things even worse than what I posted!)

 

[Flgatorguy87]

I think just to add to that, I've seen spreadsheets of draft success rates by position, I've even reproduced them here in the past, although I haven't gone looking for the source I used to quote in this post.

But the bottom line is you are going from about 30% to about 20% odds of hitting in the last 10 spots of the 1st round and early 2nd round. So you are talking about a 10% difference in odds. Think about that statistically...

1) say you pick #25, #30, #35...
Odds = 27% + 23% + 20% of getting a player = 70%

2) say you package #25+#35 to move up to #20...
Odds = 33% + 23% = 56%

There are lots of papers and studies of draft pick valuation these days, but I don't think there is any statistical argument which says trading up makes sense. It never makes sense statistically. All you can do is think you are smarter than everybody else and will beat the draft odds by getting a 17-18 year old player who in that part of the draft is probably 4-5 years from being able to play in the NHL, and thinking you are going to get the right guy, develop him right, he's not going to get hurt, all that. Which based on my experience is usually hubris. All you are going to succeed in doing is giving yourself a short-term dopamine hit for "getting your guy". In the long run, however, it is not a sound strategy.

Based on the hit rate, the chances of us NOT getting at least 1 "hit" would be 45% for 3 picks ( .73*.77*.80) and would be 51% (.67*.77) for two picks. So 55% chance of success with 3 picks and 49% chance for 2 picks.

The caveat to all of this is what is a "hit". Coin flips are easy for probabilities, but calculating what defines a success that may take 4-6 years to determine is muddy.

 

[Porter Stoutheart]

Based on the hit rate, the chances of us NOT getting at least 1 "hit" would be 45% for 3 picks ( .73*.77*.80) and would be 51% (.67*.77) for two picks. So 55% chance of success with 3 picks and 49% chance for 2 picks.

The caveat to all of this is what is a "hit". Coin flips are easy for probabilities, but calculating what defines a success that may take 4-6 years to determine is muddy.

Another good way to look at it.

There is definitely a range of "hit" outcomes between Austin Watson and Tanner Molendyk ( )

Anyway, if you really think you are that smart, go for the trade-up, I guess. I just have 40 years of having figured out that I'm not that smart behind my preference to keep all the picks.

 

[braindead]

I always think the trade down a few spots is the right move for this general reason.

And having spent time observing and being obsessed over drafts I would pick largely on skating speed and VO2 max results above all else.

 

[Weber to Josi]

If we somehow manage to leave this draft with Misa I will be absolutely ecstatic. I also think he is the most NHL ready of the top prospects in this draft.