Absolutely horrible analysis of the Stars. I can't take someone seriously that says Turco is an AHL goalie and that Alex Auld is the better netminder.
Since the lockout, Turco has had save percentages of .898, .910, .909, and .898. Auld has been .888, .880, .919, and .911.
So Auld
has been better lately, but I can't imagine feeling confident with that goaltending tandem.
Turco doesn't appear to be a very good goaltender, statistically speaking. Much to my surprise, to be honest. His performance over last season
was AHL level.
I also didn't realize we only had 8 usable forwards.
Most teams have 6-9 decent forwards who can be relied on to do tough sledding and a variety of grinders who you don't expect much out from except not get murdered by the opposition and are mostly used to give the guys up a breather. I think this is what is meant by "usable".
I also don't see how they can project Richards, Morrow, and Lehtinen to only play 25 games more collectively this season.
They base it on injury history. Morrow for example has played an average of a bit over 47 games over the last three seasons, so their estimate for about 50 makes a certain amount of sense, especially since he's not getting any younger.
Not suprised though. This is just one in a million of such prediction sites that just looks at numbers and obviously doesn't watch or pay attention to the teams.
Personally, while I have my doubts about their projection system, I'm glad to see at least one alternate vision of things that tries to look at things more objectively, rather than predictions from pundits that are based more on cliches, who made the loudest moves in the off-season, who built their team to the pundit's biases, the past reputation of teams... and, ironically, a lot by looking at how the teams did in the standings without wondering about whether they had injuries, good or bad luck, players playing over their heads or having bad runs, et caetera. They tend to not be very good nor very insightfuls.
They made statistical projections based on past performance. You can debate their statistical model (a lot), but certainly that doesn't seem to be an entirely silly way to approach things.
At the very least, it provides new ways to look at things.