Query about your decisions when to start or bench a goalie

[kevinsrangers]

i always have a personal rule never to bench a goalie unless im ahead in goalie games(not often) or i have a goalie to replace him with.( id rather let in 5 goals than bench a shutout)
ussually ill rank my goalies in my head and play my beest goalies no matter who there playing. sometimes if a better goalie is playing the top 3 ranked team and a not so good goalie is playing a bottom dweller i might take a chance but it ussually blows up in my face.
either way youll always be frustrated with goalies.

[hockeyrobot]

Well mine is definitely a combination every single option you listed haha!

But if I had to choose, in order would be:

H2H situation
This wins out every time. If I had Lundqvist and Quick so far this week, and a goalie min of 3 games, I would only play one more goalie for the week and trust that their peripherals can get me the win.

Combination of goalie season stats/L10 games
I generally won't sit a stud no matter what unless they're 2-8 in their last 10 with a 3.00+ GAA.

[BabyElias]

i always have a personal rule never to bench a goalie unless im ahead in goalie games(not often) or i have a goalie to replace him with.( id rather let in 5 goals than bench a shutout)
ussually ill rank my goalies in my head and play my beest goalies no matter who there playing. sometimes if a better goalie is playing the top 3 ranked team and a not so good goalie is playing a bottom dweller i might take a chance but it ussually blows up in my face.
either way youll always be frustrated with goalies.

Thanks for your thoughts! I haven't ever played roto, so difficult to say how my thinkings apply there.

But this is so true: "either way youll always be frustrated with goalies" LOL

[BabyElias]

Well mine is definitely a combination every single option you listed haha!

But if I had to choose, in order would be:

H2H situation
This wins out every time. If I had Lundqvist and Quick so far this week, and a goalie min of 3 games, I would only play one more goalie for the week and trust that their peripherals can get me the win.

Combination of goalie season stats/L10 games
I generally won't sit a stud no matter what unless they're 2-8 in their last 10 with a 3.00+ GAA.

Yup, thanks man! But if you have three studs and two slots, there's usually couple of benching decision in a week...

About H2H situation.
In the regular season, if I have good starts in the beginning of week but no SO, I continue playing the goalies. Greedy all the way Later in the match-up (FRI, SAT, SUN) I am starting take a look whether it is wise to bench them. I also check closely the stats and the remaining starts of the opponent team and "simulate" how much my goalie can suck before losing the peripherals, and of course vice versa.

In the playoffs it is a different game. It depends on the skater stats - if skaters are doing well it is wise to secure GAA and SV% and not shoot for the W and SO (or whatever the stats may be).

[BabyElias]

My final comments to this thread. Justin mentioned this thread in his latest article (http://www.dobberhockey.com/index.ph...man&Itemid=100) but the discussion dried up.

I came to a conclusion about this goalie start analysis and thought I share it with you. I spent a lot of time pondering about making a "Who to Start" goalie tool, but final push for the conclusion was the mistake I made in my H2H league playoffs last week. More about that in the end of this this post. This is a long post, but please take time to read this and give your comments. After this post I won't be bothering you with this thread, LOL!

First of all, to decide which goalie to start is a combination of all the things discussed in this thread. However, I have realized that it is impossible to find an empirical method for defining which one of two goalies are better start. Especially because there is also this mental factor Justin is preaching about:

But I would say 75-80% of the time, I just keep all my goalies in my active lineup, regardless of what peripheral stats might lead me to think otherwise.

Why?

Every goalie is capable of overcoming odds and coming up with wins when not many expect it.

To make this kind of tool is kind of similar as programming a computer to beat the best players in chess. Deep Blue computer won first time the reigning Chess Champion (Garry Kasparov) in 1996. And still the matches are tight between the best computers and best players. What I'm saying is that mind of a great Fantasy Manager will make better decisions by using all the factors than a dumb tool. And predicting the hockey games and goalie starts is pretty damn hard.

So, no empirical model for this. Case closed.

However, what would you say if the Frozen Pool by Jason Arbuthnot would have a "Compare-a-goalie-start" tool? You could select two match-ups of your two goalies and you would see all these factors laid down in front of you so that you can compare them easily. So similar as Game Planner tool, but it is for goalies showing the latest stats against the opponent, goals scored by opponent etc. I now have to make a lot of clicks in various sites to get all the information I need. This would save time. Should I start pitching this to mr Arbuthnot? Would you use this kind of tool it it would be available?

OK, now to the mistake I made last week. I had a close match-up in the 1st week of match-up. My goalies are Rinne, Hiller, Halak and Anderson. Opponents goalies Brodeur, Crawford and Roloson.

On Friday I had to make a decision whether to start Anderson or not. The stats at that moment were:

Mine:
GAA 1.76
SV% 0.925
Wins 3
SO 1

His:
GAA 2.35
SV% 0.911
Wins 1
SO 1

I made a decision to bench Anderson vs WAS as I figured the win is difficult to get and it was a B2B start for him. Also I was going to have good starts from Rinne vs DAL and Halak @ MIN Saturday.

OK... Anderson pitched the shutout and he was on my bench. BUT same night Brodeur also got the shutout (OTL) but luckily Roloson let 4 goals. The stats were then:

Mine (unchanged):
GAA 1.76
SV% 0.925
Wins 3
SO 1

His:
GAA 2.18
SV% 0.924
Wins 1
SO 2

Now I started worrying. He would have Saturday Brodeur @ BUF and Crawford vs ANA. Good starts. And more devastating news in Saturday evening for me was that Halak is not starting! Well, Rinne started and fared well (of course!) and both his goalies took tight losses. The stats were after this:

Mine:
GAA 1.81
SV% 0.927
Wins 4
SO 1

His:
GAA 2.14
SV% 0.925
Wins 1
SO 2

Okay, for Sunday I had still Anderson's start. My opponent dropped Roloson for Pavelec. I had to make a decision whether to start Anderson or not. I calculated (with my H2H goalie tool) that even if Pavelec has a shutout, the opponent cannot catch me in GAA. If he lets 1 goal, he needs to make 21 saves to beat me in SV%. If he lets 2 goals, he needs to make 34 saves to beat me in SV%. This assured me not to start Anderson. Well, in the end of the day it turned out Mason got the start and I won the goalie stats 3-1 (Anderson on my bench Sunday), and won also the match-up.

This was a good learning for me. I almost killed my week by benching Anderson. From now on, I think I only bench a goalie in last days of the H2H week if it is clearly looking so based on match-up situation. But when I need to make decisions over who to start, I will use all the factors mentioned in this thread. For me the most important factors are in no particular order:

- Last 5 games stats of a goalie
- the teams power rank (L10)
- the career stats against the opposing team
- match-up situation (do I need win more than peripheral stats etc)
- gut feeling

Cheers, and thank you for reading the whole post. Comments welcome!!