Wayne Smith wrote:The stats are often numerical fiction.
The tackle area is a good example of why stats do lie, of why they are completely misleading without the detail.
Super rugby coach in dismissing tackle stats shocker! Appeals to authority like this usually look better when you spray the quote across a picture of the person or someone historical(Lincoln & Hitler tend to be favourites)
But one player can make 100% of his tackles in a match and it tells you nothing about their effectiveness. Every tackle could have been a passive tackle, made behind the gain line, allowing the opposition to win quick ball without committing numbers. Another player may make only eight out of ten of his tackles, but the eight he made are dominant tackles. He made his tackles past the gain line, put the runner on his back, shut down the offload, forced the opposition to send in three or four players to clean out. And maybe two of those tackles created turnover ball off which his team scored. And maybe the two that he missed were desperate scrambling efforts. Yet the second player is ranked behind the first player on the media stat sheet. You might as well throw the paper in the bin. It would do less harm there.
This isn't what I'm talking about at all. Like, I get this stuff. Did I not point it out already? E.g Leinster having very high missed tackles and at the same the best(?) defense in the league. I was taking about the broad trend or pattern over a season. Like do you not think that something can be inferred from the disparity between the tackle stats of McGrath and Marmion?
Ruddock's tackle stats consistently too low for me to be taken seriously as a Six Nations blindside..... Ruddock's defensive stats don't stack up. - All Blacks Nil, Jan 15th, 2014
England A 8 - 14 Ireland A, 25th Jan 2014 Ruddock(c) 19/2 Tackles
paddyor wrote:Em, he was just called into the Ireland team to make his debut??
Doesnt mean he has learned to tackle.
It means Schmidt trusted him enough against a team that cut us to pieces out wide the last time we played. IMo it's statistically unlikely he was relying on tackle stats alone in making that call.
Ruddock's tackle stats consistently too low for me to be taken seriously as a Six Nations blindside..... Ruddock's defensive stats don't stack up. - All Blacks Nil, Jan 15th, 2014
England A 8 - 14 Ireland A, 25th Jan 2014 Ruddock(c) 19/2 Tackles
@wixford : just to note, yeah I am talking about my own confirmation bias. I was first to respond to paddy Re: not missing tackles because he wasn't there to make them. I think he has been a liability in defence at times. I'm not defending AB, i am surprised by the numbers though (what are Noel Ried's like, i wonder). Im defending paddy's use of the stats as a good conversation starter.
Peg Leg wrote:
The bias being his defence is bad and any error proves the point (that element of his game gets deep scrutiny). He has improved a lot this season ergo there was enough, but less so as he develops this season. To me thats the point, the focus is more on the (now less frequent) bad read than the other elements of his game. Daly on the other hand gets a pass... I've no problem with paddy raising the issue using the only available stats as a defence performance indicator.
Oh, so it's you that has the confirmation bias then! That's clear from that series of totally stat free (thank you for that anyway) judgements. I'm genuinely not sure what you mean by, "He has improved a lot this season ergo there was enough,..." typo possibly. But to contend that he has made less bad reads this season- from what was a very poor base level last season in the first place, has no basis in stats- as we have established there are none, nor either in the old-fashioned observable world. He was a definite liability in a couple of games. So to say "I've no problem with paddy raising the issue using the only available stats as a defence performance indicator." is the equivalent as saying 'I haven't a problem with someone claiming the increased number of diesel cars on the road is improving the environment'- even though we know that it's not an accurate measure at all. Same applies to the tackles equaling defence
.
i'd say you're great craic to go for pints with
Craft Beer would be Artaneboy's prefered tipple me thinks.
You know I'm going to lose,
And gambling's for fools,
But that's the way I like it baby, I don't want to live FOREVER!
paddyor wrote:Em, he was just called into the Ireland team to make his debut??
Doesnt mean he has learned to tackle.
LeRouxIsPHat wrote:Did Leo say that officially? Missed that one!
Of course he didn't but again, that doesn't make the series of poor defensive games just disappear.
LeRouxIsPHat wrote:Taking a more long term view I just feel like he was all over the place when he first came into the side and, while he still makes errors, he looks more assured now and makes fewer mistakes.
Yeah thats fair enough. Hopefully that upward curve continues.
My greatest gripe with tapatalk is the injustice and ignominy being inflicted on your avatar.
Peg Leg wrote:
The bias being his defence is bad and any error proves the point (that element of his game gets deep scrutiny). He has improved a lot this season ergo there was enough, but less so as he develops this season. To me thats the point, the focus is more on the (now less frequent) bad read than the other elements of his game. Daly on the other hand gets a pass... I've no problem with paddy raising the issue using the only available stats as a defence performance indicator.
Oh, so it's you that has the confirmation bias then! That's clear from that series of totally stat free (thank you for that anyway) judgements. I'm genuinely not sure what you mean by, "He has improved a lot this season ergo there was enough,..." typo possibly. But to contend that he has made less bad reads this season- from what was a very poor base level last season in the first place, has no basis in stats- as we have established there are none, nor either in the old-fashioned observable world. He was a definite liability in a couple of games. So to say "I've no problem with paddy raising the issue using the only available stats as a defence performance indicator." is the equivalent as saying 'I haven't a problem with someone claiming the increased number of diesel cars on the road is improving the environment'- even though we know that it's not an accurate measure at all. Same applies to the tackles equaling defence
.
Yes, my bias.
Not a typo just a nonsensical sentence. Read: he has improved a lot this season, ergo there were enough mistakes to feed my own assertion that he was very bad in defence,.....
Paddy sought to raise a topic, chose the defence stats available for AB and his peers and presented them to help form the discussion. Your contention that the stats do not present the empirical data to assess 100% of the body work therefore no stats can be included in the conversation is a bit much.
Your comparison between my statement and your diesel car line is good though, i like it. But your missing the point, i think the stats that paddy read (and likely surprised him) gave him food for thought and he posted them here to start a conversation about AB's defence. My response was to state that there were some gaps in the data, yours was to get the step ladder out for that high horse of yours, saddle up and trample all over the stats and paddy's use of them.
Im playing a longer game, I'd like Paddy to keep paying his subscription fee and continue to share some of the insights (read: stats) that the management teams also use in assessing player performance.
Whoa- so from the view on my high horse (or step ladder), your position is that if you have a hunch- cum genuine feeling on an issue, using data- no matter how inadequate or inappropriate is acceptable to attempt demonstrate it. Is that it? Poor stats, with such defenders...
"Oh, I used to be disgusted, and now I try to be amused!"
Peg Leg wrote:
The bias being his defence is bad and any error proves the point (that element of his game gets deep scrutiny). He has improved a lot this season ergo there was enough, but less so as he develops this season. To me thats the point, the focus is more on the (now less frequent) bad read than the other elements of his game. Daly on the other hand gets a pass... I've no problem with paddy raising the issue using the only available stats as a defence performance indicator.
That doesn't make sense. Errors do prove the point
That's not bias.
The focus is on because he has a habit over the last 18 months or so of doing something. It's not because of some perceived unfair bias. It's objective. It's because that's what he has been doing on the pitch consistently (including this season by the way).
He's only been a regular in the team since September 2016. Thats 14 months. Details!
Ruddock's tackle stats consistently too low for me to be taken seriously as a Six Nations blindside..... Ruddock's defensive stats don't stack up. - All Blacks Nil, Jan 15th, 2014
England A 8 - 14 Ireland A, 25th Jan 2014 Ruddock(c) 19/2 Tackles
artaneboy wrote:
Oh, so it's you that has the confirmation bias then! That's clear from that series of totally stat free (thank you for that anyway) judgements. I'm genuinely not sure what you mean by, "He has improved a lot this season ergo there was enough,..." typo possibly. But to contend that he has made less bad reads this season- from what was a very poor base level last season in the first place, has no basis in stats- as we have established there are none, nor either in the old-fashioned observable world. He was a definite liability in a couple of games. So to say "I've no problem with paddy raising the issue using the only available stats as a defence performance indicator." is the equivalent as saying 'I haven't a problem with someone claiming the increased number of diesel cars on the road is improving the environment'- even though we know that it's not an accurate measure at all. Same applies to the tackles equaling defence
.
Yes, my bias.
Not a typo just a nonsensical sentence. Read: he has improved a lot this season, ergo there were enough mistakes to feed my own assertion that he was very bad in defence,.....
Paddy sought to raise a topic, chose the defence stats available for AB and his peers and presented them to help form the discussion. Your contention that the stats do not present the empirical data to assess 100% of the body work therefore no stats can be included in the conversation is a bit much.
Your comparison between my statement and your diesel car line is good though, i like it. But your missing the point, i think the stats that paddy read (and likely surprised him) gave him food for thought and he posted them here to start a conversation about AB's defence. My response was to state that there were some gaps in the data, yours was to get the step ladder out for that high horse of yours, saddle up and trample all over the stats and paddy's use of them.
Im playing a longer game, I'd like Paddy to keep paying his subscription fee and continue to share some of the insights (read: stats) that the management teams also use in assessing player performance.
Whoa- so from the view on my high horse (or step ladder), your position is that if you have a hunch- cum genuine feeling on an issue, using data- no matter how inadequate or inappropriate is acceptable to attempt demonstrate it. Is that it? Poor stats, with such defenders...
Yep, get it out there and maybe learn something from people who know better or like to share an alternative opinion. Its a friendly space to air your inappropriate, inadequacies.
I think it was Dick Spring, or Einstein, who said “Not everything that counts can be counted, and not everything that can be counted counts.” Mind you, as we know to our cost, money can be counted and apparently it also counts.
It’s fair to discuss his defence, there are issues and we’ve seen them this year, but I’ve always felt the criticism is excessive and his mistakes fixable. For example, had Byrne allowed Nadolo to kick past him on the sideline, not complete the tackle on him (when he probably legally could have) and allow a try up his wing, well I think Byrne would have been absolutely slated. Franno would have had a column on it. That’s what Sweetnam did, but if you ask most people, Sweetnam had a good November and enhanced his reputation. That’s not to run down Sweetnam, but as peg leg says about Daly, I’d argue Byrne gets a harder time than that. Maybe that’s being judged to a higher standard as more is expected of him, or maybe he’s just got the name as a bad defender and he’ll just have to live with that forever. I dunno, but I think he’ll be a better defender for the work of the past month. Joe claims they weren’t looking back at 2015 and he suggests he didn’t want his wingers backing off, but I suspect there was plenty of studying the 2015 game and plenty of work on only bite in if you absolutely have to. Whether that will be enough to make him the best 14 in Leinster and Ireland over the next few years is hard to say just yet.
Peg Leg wrote:Im talking about confirmation bias. A trend is noted at a point in time. It is taken as a consistency at that point in time and regardless of the decrease in frequency of said trend being repeated, every instance thereafter is treated as proof of the original noted trend (regardless of the passing of time or changes made).
As simple as that is, that's also an excellent post.
LeRouxIsPHat wrote:Did Leo say that officially? Missed that one!
Of course he didn't but again, that doesn't make the series of poor defensive games just disappear.
No but unreasonably editing my post to omit the part right after that when I said the following does...
Maybe he did pay for his mistakes against Montpellier and Munster (assuming you were referring to the Glasgow game), or maybe they wanted Ferg's experience, or maybe it was a range of things. If he was dropped because of errors in those games, that doesn't change my point that he has improved overall.
Nobody is arguing that he doesn't make mistakes so don't pretend that they are.
So the use of available data points is rubbish but use of, eh, totally unbiased individual opinion is fine?
You can’t use imperfect science to make a point but I can use opinion to refute it categorically.
I think everybody agrees that he’s not a great natural defender. We can’t prove if he’s improving or not without watching every game he’s played and measuring tackles and reads and charting it to see if there’s been an improvement (volunteers welcome). Anecdotal evidence is js picking him for the squad and starting him. I’m sure the Leinster team have relevant data and that gets fed to Ireland and I’m sure Ireland do their own analysis too. That’s good enough to me and I’d pick the natural finisher with pace over McFadden.
wixfjord wrote:
Oh I know what you're talking about. But it's not confirmation bias if you don't have to go searching for false examples to confirm your bias.
neilinboston wrote:So the use of available data points is rubbish but use of, eh, totally unbiased individual opinion is fine?
You can’t use imperfect science to make a point but I can use opinion to refute it categorically.
I think everybody agrees that he’s not a great natural defender. We can’t prove if he’s improving or not without watching every game he’s played and measuring tackles and reads and charting it to see if there’s been an improvement (volunteers welcome). Anecdotal evidence is js picking him for the squad and starting him. I’m sure the Leinster team have relevant data and that gets fed to Ireland and I’m sure Ireland do their own analysis too. That’s good enough to me and I’d pick the natural finisher with pace over McFadden.
Yes, “The use of available data points is rubbish”, when it does not measure what it’s supposed to measure.
And yes again, “unbiased individual opinion is fine”- when compared to improperly used data. Data is being used here in this specific case, not in an objective manner where there’s a common acceptance of its validity, but selectively and inadequately to back up an individual opinion.
Sent from my iPhone using Tapatalk
"Oh, I used to be disgusted, and now I try to be amused!"
artaneboy wrote:
And yes again, “unbiased individual opinion is fine”- when compared to improperly used data. Data is being used here in this specific case, not in an objective manner where there’s a common acceptance of its validity, but selectively and inadequately to back up an individual opinion.
I would imagine there is a common acceptance of its validity, really you've just decided that there isn't but that doesn't make it a fact. I would take the stats with a pinch of salt and would agree that they don't tell the whole story, but that doesn't equate to them being wildly inaccurate or irrelevant, and to disregard them as you have is being selective to back up your individual opinion. It's hard to take your dismissive attitude to the stats seriously when you didn't think that a blatant missed tackle by JGP on Luke Marshall counted because "it was at close quarters and the Ulster man was felled instantly by another anyway".
neilinboston wrote:So the use of available data points is rubbish but use of, eh, totally unbiased individual opinion is fine?
You can’t use imperfect science to make a point but I can use opinion to refute it categorically.
I think everybody agrees that he’s not a great natural defender. We can’t prove if he’s improving or not without watching every game he’s played and measuring tackles and reads and charting it to see if there’s been an improvement (volunteers welcome). Anecdotal evidence is js picking him for the squad and starting him. I’m sure the Leinster team have relevant data and that gets fed to Ireland and I’m sure Ireland do their own analysis too. That’s good enough to me and I’d pick the natural finisher with pace over McFadden.
This.
Ruddock's tackle stats consistently too low for me to be taken seriously as a Six Nations blindside..... Ruddock's defensive stats don't stack up. - All Blacks Nil, Jan 15th, 2014
England A 8 - 14 Ireland A, 25th Jan 2014 Ruddock(c) 19/2 Tackles