Tech Talk: Where Does The Oilers' Rebuild Stand?

The constant talk of rebuilding in Edmonton has turned the idea of purposeful losing to ensure high draft picks into a good one in Oilers' fans minds. The local and national media has declared the rebuild a smashing success. The talk now is about arrows and directions and timelines.

But how do we know if the rebuild is truly going well? What do we compare them to and how do we measure progress? How do we know if the plan that was initially in place was a good one?

JPinVA at Lighthouse Hockey had a very good idea, namely applying systems engineering methodology to the Islanders' rebuild in his article "System Life Cycle: In Which Stage is the New York Islanders Rebuild?" Because I'm a software guy at heart, I'm going to steal his idea and compare the Oilers rebuild to the Software Development Life Cycle (SDLC).

Initiation: Customer or internal sponsor requests a proposal.

In the Oilers' project, this never happened. There was no internal management request to rebuild.

Concept Development: Defines the initial scope of the request. A feasibility study is performed which includes a cost-benefit analysis and a risk management analysis and mitigation plan.

Again, in the Oilers' project, this didn't happen either. There's a push by some media outlets and a fan segment to advance a narrative that the rebuild was Steve Tambellini's idea from the jump. That isn't the case. Burning the building down and starting over with Taylor Hall was NOT planned. The season before finishing last, Steve Tambellini signed a veteran coach known for grit and playoff-type hockey. He signed Nikolai Khabibulin, an ancient goaltender to a 4-year guaranteed cap hit of $3.75 million per year. He was prepared to send Dustin Penner, Andrew Cogliano and Ladislav Smid to Ottawa for Dany Heatley. After all was said and done, he spent to the salary cap.

Those are not the actions of a man defining the early scope of his rebuild.

For a man who spent so much time assessing (about 18 months, give or take a month), it's astounding that this step was skipped.

Planning: Completion of the project management plan, which includes a resource plan, defining the necessary resources to achieve a successful implementation.

This is one area where Steve Tambellini partially completed the phase. Tambellini, unlike his predecessor Kevin Lowe, recognized that a 30th-place AHL team and a 30-rate organization probably wasn't the best way to build a winner. Granted, it doesn't take Isaac Newton to figure something like this out, but given Edmonton's prior lack of planning, it is a plus. However, given Tambellini's signings and trades at the NHL level, it's probably fair to say that he didn't finish the planning stage, unless the resource plan consists of nothing but players acquired through the draft. This, however, is a flawed plan.

Requirements: Uses the documentation developed in the Concept Development and the user needs to create a detailed functional requirements specification or document.

The requirements here should be simple: build a winning hockey team. But what does that winning hockey team consist of? The Oilers have shopped the Detroit model, the Chicago model and the Pittsburgh model to the media, but all three teams were built quite differently, and none of those models included an over-the-hill goaltender making 4 times too much money backstopping them to the cup. Well, Tallon's original plan included that goaltender, but he was jettisoned, then they won the Cup. There are a couple of things each team had in common: depth up front, a strong top four, a couple of shut-down guys to play the penalty kill and two-way centers. Pittsburgh and Chicago also had insanely hot goalies at the right time in the playoffs, but planning for that is foolhardy. It would cause a GM to sign a poor contract with a goalten....

Design: Uses the functional requirements to create a detailed design document.

Kevin Lowe has publicly stated the Oilers are in the midst of a six-year plan. Couple that with "build a winning hockey team" and the lack of NHL talent acquired under Tambellini's watch and the design starts to take shape. 18 months of assessment and repeated assurances of "clarity" created the design document.

Development: Executes the steps necessary to create application software from the design document. Preparation, coding, compiling, unit testing and bug fixes are all part of this phase.

It's quite dangerous to initiate the development phase of a project without a passing through the first five phases, but that's what the Oilers have done. They've rushed some players into roles that they cannot possibly handle and been lax to try and find players who can handle those roles. Some players have shined, others have faltered. Except for Eric Belanger, there were no players added in the off-season who can be expected to handle the tougher roles, so the cycle will likely happen again in 2011-12.

At the minor league level, the Oilers are doing a bit better. Mike Sillinger is talking to the prospects in the organization, another new trend in Edmonton. The AHL team is giving minutes to players in a controlled environment, covered by proven AHL veterans with some miles on them. It gets a little tougher this season though because the Barons are about to be overrun by rookies. If Coach Nelson manages to navigate a season with so many rookies and keep the team competitive, the Oilers might be in good shape. If the kids are thrust into roles they cannot handle, things could go south for the core of the 2nd-5th round picks.

As in any software company that jumps to this phase too early, management now has to hope the engineers themselves can find their way on their own, because there isn't much in the way of requirements or process to guide them through to the end.

Test: Quality Assurance performs full system testing and integration testing to ensure final product meets the specifications in the functional requirements document.

Some fans will argue that this year is the first big regression test - Tambellini is throwing out some compiled, but not unit-tested code and he's going to see what he has and go from there by overwhelming the QA department. It goes back to the development cycle. Throwing people into roles that they cannot handle isn't good management, but if the players themselves handle it, managers will look like geniuses. Essentially, if this is one big test, the success of the project depends solely on some lucky code getting through QA and into the field.

Implementation: Deployment of the final product into a live environment.

If this is the final product, there are serious issues with it and at this point, it's got more bugs than features.

Maintenance: Ongoing support of the product in a live production environment.

I think it's safe to say that the Oilers would require the world's largest support organization if this were the final product.