Where did your figure of 21% come from? According to the stats I use (OPTA) the figure is 16% (51 long passes out of a total of 309).
OPTA's entire business depends on their being accurate! Regarding the Newcastle match, if carrabuh's figure were correct it would indeed represent an unusually high long pass count for us; 16% though is only marginally above our average for the current season (15%). To put this in perspective, the average (and median) for the PL so far this season is 13.5%. Four teams have averages of 9-10% (Arsenal, Liverpool, Man City, Swansea), while at the other end of the scale three teams average 17-18% (Villa, West Ham, Palace). We are in a group of 4 clubs on 15% (along with Sunderland, Hull and Fulham); Stoke and WBA make marginally less use of the long ball (14%), Cardiff marginally greater use (16%). The 4 teams on 9-10% are often (rather ludicrously) described as "passing teams", (i.e. short passing/possession-keeping teams); Chelsea, Everton, Man Utd, Newcastle, Spurs and Southampton form a group whose use of long passes is just below the average; the three at the top end (Villa, West Ham, Palace) are the real "long ball teams" in the division. The group which includes us (i.e. those who make slightly above average use of long passes (14-15% against the average of 13.5%) are not IMO correctly described as long ball teams. They are teams which aspire to play a more short passing game, but currently lack the players to do it. In particular, they are wary of short passing their way out of defence when under pressure.
