But If You Plug In Phony Bullshit Numbers, It All OK!

Looks like the hypocrosy is running wide and deep @ the NHLPA. After claiming for months that the NHL’s numbers are “fundamentally flawed”, they claim to have pluged those very same numbers into the NHL’s forcast projects and “proved” a difference of nearly a billion dollars, from losing $569 million over three years to making $412 million. Wow, if that’s true, that’s some pretty shitty math by the NHL. Of course, it’s complete bullshit.

The NHLPA’s argument rests on the projected revenue and player cost growth for the next three years. The NHL projects that revenue will rise at 3% per year while the players claim the NHL’s own historical numbers suggest that revenue will rise 7.8% per year. And it’s true that over the past five years, the NHL’s Y/Y revenue growth has been 7.8%. Of course, during that time, there has been three new teams added as well as TV contracts that have come and gone (the current NHL TV contract w/ NBC is worth much less than the previous one with Fox). If you look at just the past four years – a much more relevant time frame – revenue has gone up between 4.3% and 6.4% per year. Given the effect of the lockout, I’m thinking 3% is a pretty good estimate.

The NHLPA goes on to argue that the NHL’s own numbers suggest that player salaries will only go up 7.3% rather than the 12% the league suggests. Again, the 7.3% number for five years is accurate, but if you at the player cost growth for 01-02 and 02-03 the increase was 11.6% & 11.8% Y/Y. (03-04 was a paltry 1.8% Y/Y increase, but I would attribute that to the looming lockout.) What’s really scary is that in 01-02 and 02-03, total player costs went up more per year than in the two years prior when new teams joining increased the total number of player jobs! In 00-01, when two new teams (around 50 new players) joined the league, total player cost went up $121 million. The following two years, when no new teams were added, total player cost went up $127 and $144 million respectively. Gee, looks like the 12% estimate is pretty close to reality too. Plus, the league admitted that the player cost growth might drop as low as 9% based on the 24% salary rollback and other deflators the players proposed. But last I checked 3% revenue growth + 9% player cost growth = bad news for the owners in the long run.

I’m thinking the NHLPA has a bunch of excel users punching in random historical data into the model to see what provides the best outcome for the league. In their “rebuttal”, they keep switching their historical model - first they use 10 year historical revenue growth numbers, then they use five year historical revenue and player cost growth numbers. Then, in their final chart uses an unprovided player cost growth estimate that is actually marginally higher than what the league is projecting, cutting the difference from the NHL’s projection in half to $440 million.

If they can’t even keep their story straight on their web page, do they really expect to fool anyone else with this Enron-esque number crunching?

More Transparency…But I’d Like To See More Still

I saw a post on TSS.NET pointing to the Outlook 2003 Integration APIs:

Microsoft has released former “internal only” APIs for developing applications that work with Microsoft Outlook. The APIs include account management, connection state, free/busy, MAPI-MIME conversion, and the Outlook store.

I have no idea why these APIs have been released at this time, but providing more options/control/tools to developers is better than less – along the same lines of the ASP.NET 1.1 Membership Prototype.

My only issue is that not all the APIs are documented. For example, recently I was interested in being able to progammatically set Outlook’s online/offline status. When I read the post above, I thought the Connection State API would be the answer. Alas, that API can only be used to get notification when the connection state changes. Or, at least, that’s all that’s documented: the IMAPIOfflineMgr interface has seven undocumented methods and the IMAPIOffline interface has two. I’m sure there’s a good reason those others aren’t documented, but it’s frustrating all the same.

Of course, even with the ommisions, that’s a lot of cool stuff to add to Niobe. Simon?