He is sheltered getting 65+% offensive Zone starts and being strapped to a very good defender in Carlo 5v5. That is what it took to bring his stats up to respectable. That means without heavy sheltering, he would be garbage. Who are we strapping to him to help make up for his defensive deficiencies? Parayko? Ok, then who is getting the defensive zone starts? Right. We do not have a Carlo and a McAvoy to carry Krug's GARBAGE defensive zone play. We have Parayko and..that's it. Parayko can either cover for Krug or carry the defensive weight in the toughest situations. He can't do both.
Edit: Even you are saying we need to build the rest of the defense to cover for his defensive deficiencies. How is that not garbage defensive play when it takes 5 other guys to cover for how bad 1 guy is? Of course you can build a good D with one guy who is garbage at D but really good at offense (elite would put up more than 60 points given the PP time on that PP and the O-zone starts). But we don't have that D group to do it. We have too many other guys who also need to be sheltered to make it remotely possible to build that type of D in any easy or quick manner. As constructed his garbage defense is going to hurt us a lot more than his good offense will help us. That isn't even getting in to how our forwards are not constructed to take advantage of his quick strike skills.
This is just not even remotely true. His stats were better before he was played with Carlo regularly.
In 2013/14 he was a +18 and he played 200+ minutes with each of Millar, McQuaid and Boychuck. He had a GF% of 59% and an expected GF% of 55%. His offensive zone start percentage was 65%.
In 2014/15 he was a +13 and he played 200+ minutes with each of Millar, McQuiad and Seidenberg. He had a GF% of 58 and an expected GF% of 53%. His offensive zone start percentage was 60%.
In 2015/16 he was a +9 and he only played 200+ minutes with each of Millar and McQuiad. He had a GF% of 56% and an expected GF of 49%. His offensive zone start percentage was 53%.
In 2016/17 he was -10 and was partnered almost exclusively with McQuiad. He had a GF% of 47% and an expected GF% of 57%. Interestingly enough, this was the first season he ever played with Carlo. They played 130 minutes together and had terrible outcomes even though they had good underlying metrics. They allowed 7 goals while scoring only 1, despite starting in the offensive zone 70% of the time and having an expected GF% of 57%. In his time paired with McQuiad, Krug saw a GF% of 47% with an expected GF% of 55% and starting in the offensive zone 55% of the time. In year 1, Krug's numbers took a noticeable hit when paired with Carlo.
In 2017/18 he was an even 0 and played 200+ minutes with Carlo and McQuaid (he played with Carlo for triple the minutes though). He had a GF% of 53% and an expected GF% of 55%. He started 65% of shits in the offensive zone.
In 2018/19 he was -2 and Carlo was the only guy he played 200+ minutes with (although he had 100+ minutes with 3 other guys and another 85 minutes with McAvoy). He had a GF% of of 58% and an expected GF % of 54%. He started 66% of shifts in the offensive zone. His GF% was higher during the 222 minutes he played with Connor Clifton and Kevan Miller than it was when he was paired with Carlo (although it was 59% or better with all 3). It was slightly worse in the 85 minutes he played with McAvoy (57%). He was awful when paired with Moore.
Last year he was -4 and again played the majority of his minutes with Carlo. However, in the 175 minutes he played with McAvoy he had a slightly better GF% (60% to 57%) and a substantially better expected GF% (62% to 49%) than he did when paired with Carlo.
His underlying numbers are worse with Carlo than they were before Carlo broke into the organization. His underlying numbers for the last couple years have been worse with Carlo than they have been when he is paired with other D men on the Bruins. There is absolutely nothing true about the claim that Carlo has dragging him around the ice or that his underlying metrics were poor before they got a guy like Carlo to prop him up. If anything, there is evidence to the contrary.
Krug absolutely needs to be sheltered with zone usage because his defense is not good. But his defense isn't even remotely as bad as you make it out to be. Throughout his career he has consistently helped score more or break even while he is on the ice at even strength, regardless of who he is paired with.
I have been very vocal that I don't think $6.5M is a good use of our cap space for him because our blueline isn't built properly to support what he brings to the table. I agree with you that he isn't good defensively, but he isn't close to the black hole at 5 on 5 you believe he is. There is a huge range of data points between "garbage who is a complete liability" and "average defensively." Krug falls between those two points and is closer to the latter than the former.
Frankly, the bulk of his underlying numbers suggest that him and Dunn are extraordinarily similar defensively at even strength. They need to be sheltered in terms of usage, but can be effective defensively in that role so long as their partner is a replacement level defensive minded D man. I think Krug would do very well at 5 on 5 with a guy like Bortuzzo.