Assorted links


As per the couch... Just insert 1/4" AR500 steel into the back of any off the shelf couch. You may have to reinforce the couch (and your floor) because of the weight but it'll stop anything that can be realistically carried and shot by a human or non-cyborg.

If you need a bulletproof couch, I'd consider that a pretty serious Feng Shui Fail.

I think I saw this couch in Mr. and Mrs. Smith. Or was it a refrigerator?

I've seen some people apply statistics to football (soccer), and I know that the statistical methods tried in American baseball are beginning to be applied in Europe, but other than that football has to be one of the sports where statistics are invoked the least. For example, when Barcelona win a game 7-0 they're the best team in the world again, but when, in the next match, they draw 1-1 there is a major malfunction in the team. Apparently, few fans and sports commentators know about regression to the mean.

The retirement of Sir Odious Ferguson from Manchester United is a good thing in itself, irrespective of the club's results this season.

What's also interesting is that of the few stats they do have, they can have pretty ambiguous meaning. For example, one of the most frequently cited is "shots on goal". Now, is higher better, implying good control of the ball on the opponents side of the pitch? Or is lower better, implying better accuracy, and more calculating and effective shot control? In Basketball, high shots on goal is generally viewed as negative, but I've never gotten the impression of such an attitude from (soccer) commentators, who will quote the number, but never comment on it.

High shots on goal is a good thing. No idea why high "shots on goal" would be a negative in basketball either.

High shots on goal means you are missing a lot.

Shots on goal represents scoring opportunities. In basketball, high shots on goal is bad because the net is undefended from the shot and a miss is more likely to result in a turnover. In hockey and soccer, the goal is actively defended, so shots on goal substitute for accuracy. The relative scores of basketball and hockey/soccer says it all.

I've never heard of shots on goals as a statistic used in basketball. Field goal percentage, points per shot, etc. all of the time, but never shots on goal. Did you mean to write handball? Not meaning to be critical, just curious, as I have no idea of how people outside of North America understand basketball.

As you imply, for a given score, shot percentage is interchangeable with shots on goal. It is just that in soccer, that is how it is reported. Partly, this is because of the low scores in soccer, partly because of the abysmal percentage, and partly because in basketball FG percentage is broken down by player, i.e., the numerator isn't easy to decipher based on the final score.

A big part of this though is that, in soccer, high shots on goal being good or bad depends on the tactics of a given game (or stage of a game). Sometimes, an offense will pass among themselves, control the ball, and look for an opening to take a shot, and a missed shot will frequently result in giving up possessition (not unlike most of basketball). Sometimes this is clock management, but not always. In other situations, a team's primary tactic will be to keep up pressure on a defense, taking frequent shots on goal with the hope that the defense will overcommit, opening up a shot for a rebounder.

The point being, shots on goal is a very ambiguous stat in soccer, and that is one of the very few stats typically reported in games.

To answer your query though, I'm American. I was just trying to keep my original comment succinct by using with the terminology I've seen in soccer commentary, overseas.

Nobody who knows anything about basketball thinks that more field goal attempts (that is what people who actually follow basketball call "high shots on goal") is a bad thing.

I was puzzled by this in the article and now I'm puzzled again by your reference to it. How is there regression to the mean? What mean? The New York Yankees' World Series and pennant records will revert to average in the long run? The Vikings and Cubs will eventually have an average number of league championship victories? The US, Russia, and China will some day not dominate the Olympic games, and Ireland, Botswana, and Ecuador will be medal count contenders?

Moneyball is a fluke strategy that can be neutralized by money. Winning games or playing in them is not a memoryless process.

How stable are things like batting averages?

Regression to the local mean. Performance = talent + luck. When a team performs very poorly, the posterior probability that their luck was poor is high (albeit not certain). Thus, given that luck is mean zero by definition, the probability that the liuck component is higher than the mean luck in local history is high. That's how the article example is quite different than than your Yankees-Cubs examples, since is those talent is not even quasi-fixed in the longer run, as it is with a season.

Batting averages are pretty stable, though not as stable as some other offensive metrics.

I've heard of "security theater" (thanks to MR) but what about "team coach theater". I have always suspected the role of the coach in sports like soccer, where there are no time outs, was over-rated

World's first? I'm pretty sure bulletproof couches existed long before now, at least when people in movies need to hide from a bad guy.

+1 Along with bulletproof car doors.

re #2: further evidence for the clear superiority of American football. No other sport has the wonderful blend of tactics vs strategy, requiring both team work and inspiring individual athletes, and clear impact of coaching (through direct on-field play-calling). Enough with this soccer fad already :)

#1: Should be renamed to "That culture that is 'merica"

#2 i think this is slightly disengenuous; the study conclusions are reasonable (and have been discussed at length alongside the similar University of Muenster study), but obvious and completely and utterly beside the point.

In the circumstances studied, the line between success/failure is ridiculously slim. Clubs can pick from a small pool of potential employees, with very similar skillsets. This employee (or the clubs director of football) then can pick a from a slightly larger set of similarly-skilled players (and from a tiny, tiny pool of players who are both mentally and technically ideal). All this in an attempt to make minute improvements that could, possibly, translate into vast monetary gains and glory. Averaged out across clubs in a division, there'll obviously be very little change. But that's not the aim of the game, the aim is to be an outlier.

Also, a manager a) is in direct control of a small group of athletes - if that group, as whole, despises said manager, and this is reflected in performances, replacing that manager is a completely viable strategy, and b) the manager is a lighting rod for the frustrations of fans - if those frustrations boil over, and continue to boil over, removing the manager to reduce pressure on players (and owners) is again viable. Sure, if things were to be left to fester, a few good results might average their way through, but at the cost of long-term damage. The study also doesn't take into account that different managers may be useful for acquisition of [perceived] higher-calibre players (Pocchetino and Osvaldo? DiCanio and Giaccherini?). Or to change a culture at the club (eg DiCanio again vs. piss poor player discipline/professionalism at Sunderland). And then, for clubs with money + high-quality players, there is Chelsea's short-term replacement strategy - hire a firefighter with an excellent reputation until a high-calibre manager can be tested long-term.

* edit: the way this has been portrayed in the article, and in previous articles over the last 2 years, is slightly disengenuous

Comments for this post are closed