Billy Barlow

William McKenzie "Billy" Barlow was an important part of the mighty Montreal AAA teams of the early 1890s. Born in 1871 in Montreal, Barlow began his senior career playing point with the Montreal Victorias, but soon joined the Winged Wheelers, switching to centre where he made his impact on the game.

Barlow was never the best player in the game, but he was really very good, and a conscientious all-rounder at that. He was known for his quick, dodging runs, and excellent combination play with rovers Clare McKerrow and Havilland Routh - an excellent player who will be profiled here in time. Above all, Barlow was noted for his blazing speed.

Barlow was the bright particular star all through. He skated faster and played better than any other man on the ice... (Montreal Gazette, 18 Jan 1897)

Barlow came around the flags with the velocity of a rocket and was soon half-way down the ice, with no one to intercept him. (Montreal Gazette, 22 Feb 1897)

But Barlow did not simply use his skating to start rushes, he was also diligent in his forechecking. For example:

From the face the rubber went to the Ottawa end; Barlow followed it closely, [Baldy] Spittal fumbled, Barlow went behind the flags and pushed the puck forward to [Norman] Dawes, who touched sufficiently hard to score. (Montreal Gazette, 27 Feb 1897)

Great speed, good goal-scoring ability, excellent playmaking, and aggressive forechecking. What more could you want in a forward?

Barlow...was the star. He played his usually brilliant game, taking advantage of everything. (Montreal Gazette, 1 Mar 1897)

Season	Club	League	Pos	Age	GP	MIN	OP	DP	PP	TPA	TPAK
1891	Montreal Victorias	AHAC	1	19	20	500	-0.2	0.7	0.0	0.5	1.00
1892	Montreal Winged Wheelers	AHAC	15	20	32	720	-0.9	-0.4	0.0	-1.3	-1.81
1893	Montreal Winged Wheelers	AHAC	5	21	70	1400	4.8	1.4	0.0	6.2	4.43
1894	Montreal Winged Wheelers	AHAC	5	22	80	1600	5.5	1.6	0.0	7.1	4.44
1895	Montreal Winged Wheelers	AHAC	5	23	80	1600	1.3	1.1	0.0	2.4	1.50
1896	Montreal Winged Wheelers	AHAC	5	24	80	1600	3.4	0.2	0.0	3.6	2.25
1897	Montreal Winged Wheelers	AHAC	5	25	80	1600	5.4	1.7	0.0	7.1	4.44
					442	9020	19.3	6.3	0.0	25.6	2.84

Art Duncan - Scoring Champion from the Defence?

You may know that Art Duncan is credited as being only one of two defencemen to lead a major professional league in scoring, the other being (of course) Bobby Orr, who did it twice in the 1970s. In 1923/24, skating for the Vancouver Maroons of the PCHA, Duncan played 30 games, scoring 21 goals, 10 assists and 31 points, all of which were tops in the league. So he did something even Orr wasn't able to: as a defenceman, he not only led the league in points but in goals scored.

Or did he?

Are there reasons to be suspicious of this claim? The fact is, although players are generally referred to as having played only their primary position, players from Duncan's era and earlier often moved around a lot through their career. Just because Duncan is listed as "Defence" doesn't mean he only ever played defence. Have a look at career major-league scoring record:

Season	Club	League	GP	G	A	TP	PIM
1916	Vancouver Millionaires	PCHA	17	7	4	11	25
1917	Toronto Northern Fusiliers	NHA	6	4	1	5	12
1919	Vancouver Millionaires	PCHA	18	2	1	3	0
1920	Vancouver Millionaires	PCHA	18	4	8	12	0
1921	Vancouver Millionaires	PCHA	24	3	5	8	6
1922	Vancouver Millionaires	PCHA	24	5	9	14	25
1923	Vancouver Maroons	PCHA	26	15	6	21	8
1924	Vancouver Maroons	PCHA	30	21	10	31	20
1925	Vancouver Maroons	WCHL	26	5	5	10	28
1926	Calgary Tigers	WHL	29	9	4	13	30
1927	Detroit Cougars	NHL	34	3	2	5	26
1928	Toronto Maple Leafs	NHL	43	7	5	12	97
1929	Toronto Maple Leafs	NHL	39	4	4	8	53
1930	Toronto Maple Leafs	NHL	38	4	5	9	49
1931	Toronto Maple Leafs	NHL	2	0	0	0	0

The 1923 and 1924 seasons certainly seem to stick out. Is it possible that he played forward in those seasons? His teammates might provide a clue. In 1923, for defenders Vancouver had Lloyd Cook (who himself played a bit of forward in his career), and...that's about it. Abbie Newell, another forward/defenceman who played only six games, was on the team as well. Frank Patrick played only two games. Everyone else was a forward throughout their careers, suggesting that Duncan must have been a defenceman in 1923. As for his high scoring total, he was actually outscored by Cook by 30 points to 21. Both of them were basically on the ice 60 minutes per game, while the forwards were subbed out fairly often; Mickey MacKay led the team with 40 points, and Frank Boucher was next among forwards with 20.

But the following season, the Maroons added Joe Matte and Helge Bostrom to the lineup, both of whom played both defence and forward in their careers, and both of whom played regularly in 1924. Along with Duncan and Cook, this gave them four possible defencemen. Duncan scored 31 points, Matte scored 15 and Cook 12. Bostrom was clearly a sub, with only three points in 26 games. But perhaps Matte and Cook were the defencemen, Bostrom a substitute for them, and Duncan was on the forward line.

Of course, we don't really need this kind of analysis, although sometimes it can provide some insight. We can actually just go to the game reports published in the newspapers at the time, and see what position Duncan was listed as playing. So as best as I can figure them, these are the positions for the 1923 Vancouver Maroons, as they were listed in the starting lineups (SD is for subsitute defenceman, and SF substitute forward):

	G	RD	LD	C	RW	LW	SD	SF	Total
LEHMAN, Hugh	25	0	0	0	0	0	0	0	25
REID, Charlie	5	0	0	0	0	0	0	0	5
DUNCAN, Art	0	24	0	0	0	1	1	0	26
PATRICK, Frank	0	1	0	0	0	0	1	0	2
COOK, Lloyd	0	0	30	0	0	0	0	0	30
MacKAY, Mickey	0	3	0	22	0	3	0	2	30
DENNENY, Corb	0	0	0	2	1	0	0	18	21
TAYLOR, Fred	0	0	0	1	0	0	0	0	1
SKINNER, Alf	0	0	0	0	17	0	0	6	23
HARRIS, Fred	0	2	0	0	8	6	0	4	20
PARKES, Ernie	0	0	0	0	2	0	0	27	29
BOUCHER, Frank	0	0	0	5	2	17	0	5	29
ASSELSTINE, Jack	0	0	0	0	0	2	0	10	12
NEWELL, Abbie	0	0	0	0	0	1	1	4	6
COTCH, Charlie	0	0	0	0	0	0	0	15	15
	30	30	30	30	30	30	3	91	274

Duncan was the starting right defenceman in 80% of Vancouver's games. But he did start once at left wing; maybe that's an indication that the following season he played more and more forward, allowing him to score at an unprecedented rate.

Well, in 1924 the Vancouver Maroons played thusly, as near as I can figure it:

	G	RD	LD	C	RW	LW	SD	SF	Total
LEHMAN, Hugh	30	0	0	0	0	0	0	0	30
DUNCAN, Art	0	30	0	0	0	0	0	0	30
COOK, Lloyd	0	0	27	0	0	0	1	0	28
PATRICK, Frank	0	0	1	0	0	0	3	0	4
BOSTROM, Helge	0	0	0	0	1	1	17	7	26
MATTE, Joe	0	0	2	0	0	5	15	7	29
BOUCHER, Frank	0	0	0	29	0	0	0	0	29
COTCH, Charlie	0	0	0	1	0	1	0	12	14
SKINNER, Alf	0	0	0	0	27	0	0	2	29
PARKES, Ernie	0	0	0	0	2	1	0	27	30
MacKAY, Mickey	0	0	0	0	0	22	0	6	28
	30	30	30	30	30	30	36	61	277

So no, Art Duncan did not play forward, at least as far we can tell. He started on defence every game that season. Just as Lloyd Cook put up impressive scoring numbers the season before while playing defence, Duncan did the same, but took it to a whole 'nother level. It seems that Duncan simply scored at a torrential pace, and deserves full credit for the accomplishment.

But there's still something else to consider. In 1922/23 and 1923/24, the PCHA played an interlocking schedule with Canada's other western major league, the WCHL. In 1924, for example, Vancouver played 11 games against each of Victoria and Seattle, the other PCHA clubs, and two games against each of Calgary, Edmonton, Regina and Saskatoon of the WCHL. Although Vancouver was technically in the PCHA, they played 27% of their regular-season matches against WCHL clubs. For all intents and purposes, the PCHA and WCHL were two conferences of the same league, a league with an imbalanced schedule. And just as we don't give Henrik Sedin a 2012 scoring championship because he led the Western Conference in scoring, we shouldn't be looking at interlocking leagues as being separate.

This is especially important in this case, because it's clear that the WCHL was the higher-quality league in 1924. In the "inter-league" games, the WCHL teams had a combined 17-5-2 record. This means that Art Duncan had lesser competition to beat for the scoring championship of his league. If we combine the WCHL and PCHA scoring leader lists, we get the following:

Player	Club	Pos	GP	G	A	TP
COOK, Bill	Saskatoon Crescents	7	30	26	14	40
OLIVER, Harry	Calgary Tigers	5	27	22	12	34
HAY, George	Regina Capitals	6	25	20	11	31
KEATS, Duke	Edmonton Eskimos	5	29	19	12	31
DUNCAN, Art	Vancouver Maroons	3	30	21	10	31

Bill Cook of the Saskatoon Crescents led the WCHL/PCHA in goals, assists and points. He outscored Art Duncan by five goals and four assists, or nine points, almost 30% more than Duncan. Crediting Art Duncan with a scoring championship is silly; he technically led the PCHA in scoring, but the PCHA was not a self-contained league at the time. Duncan only wins the scoring title if you ignore two-thirds of the teams that he played against in 1924.

So while Art Duncan was certainly a defenceman when he scored 31 points in 1924, he should not be given credit for a scoring title. The PCHA and WCHL were separate leagues in name only; since they played against each other they were simply conferences of the same effective league. Bill Cook led this league in all offensive categories. Duncan's numbers were certainly very impressive, but Bobby Orr is alone in leading a major professional league in scoring from the blueline.

Hockey's First Hat Trick

You might know that the first recorded, organized indoor hockey game was played at the Victoria Skating Rink in Montreal on March 3, 1875. You might also know that less than two weeks later, on March 16, the first such game played between two clubs; the March 3 match was made up of two sides of Victoria rink skaters while March 16 featured the Victorias versus the Montreal Football Club.

But do you know who scored hockey's first recorded hat trick? It happened on February 26, 1877. There are at least three matches before that date that might have featured a hat trick, but we don't know who scored all the goals in these games, so we have to satisfy ourselves with the first such recorded event. Who accomplished the feat? Would you believe a 37-year-old stockbroker?

Charles Geddes was involved in the creation of the Montreal Stock Exchange in 1874. But more importantly (for our purposes, anyway), he played hockey with James Creighton, the father of the organized game, who captained many a side and played back while coaching his players.

Creighton was captaining the Metropolitan club against the St. James squad on that February day. The Mets won the match 3 games to nil, Geddes taking each goal himself. The Montreal Gazette covered the game in some detail in their February 27, 1877 edition. Geddes' goals are described thusly:

"...Geddes soon came to the rescue, and toying with the ball in his particular and graceful way, succeeded in putting it through the goal and scoring the first game for the Metropolitans. Time, 15 minutes."

"...unfortunately for the St. James' men, Creighton made some admirable play into Geddes' hands, who put the ball through in a twinkling, thus scoring the second game for the Metropolitans. Time, 3 minutes."

"...and Geddes again took the ball...and sent it through the goal for the third time, making the third game for his side. Time, 7 minutes."

So Charles Geddes can be considered hockey's first scoring star. But, this game was also the last recorded game we have of him playing. In the 1870s we only have records of a few matches per season, and with a player of this age it's probably not surprising.

Side Note: The description of the second goal above suggests that James Creighton, besides being father of the game, should be credited with hockey's first recorded assist. Not bad considering he was a defenceman.

Credit goes to hockey historians and SIHR members Pat Houda and Carl Giden for the source of this information.

An Even Earlier Stick

In the last post we had a quick look at the evolution of the early hockey stick, and the effect it may have had on scoring in the game. Today I give you a little bonus. I found a high-quality image of a stick from c.1880, five years earlier than the oldest one poorly illustrated in the last post. The 1880 stick is illustrated below, along with the 1885 and 1893 editions of the implement.

It looks like the first big change was in fact the shape of the blade, which went from a rounded shape (much like an early field hockey stick blade but with more surface area) to something more oblong in the early 1880s, which greatly resembles a modern blade.

Early Evolution of the Hockey Stick

We recently looked at the increase in hockey goal-scoring from 1892 to 1911, and hypothesized a reason for it. I wanted to go back a bit now, to look at another, rather more dramatic, increase in goal-scoring over time. I'll posit a hypothesis to explain this one as well.

First, to the facts. Organized hockey tournaments began with the 1883 Montreal Winter Carnival, which was won by McGill, and also featured only 1.13 goals scored per team per game. Such low scoring rates continued in the 1884 and 1885 Winter Carnivals, and the 1886 Montreal hockey championship (held due to the lack of a Carnival that year). In 1887, between the Winter Carnival games and challenge matches of the newly-formed Amateur Hockey Association of Canada (AHAC), over two goals were scored per game, and starting in 1888 is was substantially higher than that, generally approaching three per game. The data is here:

Season	GPG
1883	1.13
1884	1.43
1885	1.17
1886	1.18
1887	2.09
1888	2.88
1889	2.79
1890	2.93
1891	2.50
1892	3.14
1893	3.88
1894	3.03

Something clearly changed around 1887, and as well already know, it's nothing to do with the roles players played on the ice. My hypothesis is that it is largely to do with something that can be seen by comparing portraits of the Montreal AAA hockey teams from 1885 and 1893. First the 1885 club:

And now the 1893 players. Take note of the long wooden items in their hands.

The sticks are noticeably longer. The 1885 sticks seem more like field hockey sticks, and required players to hunch over quite a bit when stickhandling. The 1893 sticks, while still short by modern standards, are a fair  bit longer and the blades more closely resemble modern ones as well. I believe that the additional length of stick made stickhandling and shooting that much more accurate, and as such scoring increased all around.

By spending some time with some old hockey photos and a ruler, I came up with some estimates of the length of hockey sticks at various points in early history. We know that for the 1927/28 season, the NHL instituted a rule that sticks would be no longer than 53 inches, measured from the top of the handle to the heel of the blade. In earlier years, sticks were shorter than this, apparently increasing in length fairly steadily over time, except for a noticeable jump somewhere between 1885 and 1893.

In 1885, hockey sticks were about 33 to 35 inches, measured from the top of the handle to the heel of the blade.

In 1893, they were about 42 to 46 inches.

In 1901, they were about 44 to 49 inches.

In 1910, they were about 46 to 52 inches.

As an illustration of these differences, here are some badly-drawn hockey stick shapes:

So it seems to me that hockey sticks likely became substantially longer around 1887, and this in turn led to dramatically increased scoring as players were provided a much greater ability to control the puck, both on a rush and in a shot. Competing ideas are welcomed.

Offensive Contribution from Defencemen

Recently we had a look at the evolution of positions in hockey. We noted that from about 1892 to 1911, the standard lineup could be called 1-1-1-4, being a goaltender, a point, a cover and four forwards. Even though this formation persisted for about 20 years, the results on the ice varied quite a bit over this time. In 1894, AHAC teams scored an average of 3.03 goals per game. This number trended upward, and spiked around 1904, reaching a high of 7.73 goals per game in the ECAHA in 1907.

Obviously there must be some reason for this, and looking for some answers should be interesting. I've got one hypothesis already: as time went on, the defensive players (point and cover-point) became more and more involved in the offence. As the seasons progressed, the proportion of goals scored by points and cover-points increased largely in step with the increase in goals-per-game. Note that I don't mean that as teams scored more goals, defencemen scored more goals in step with the forwards; I mean that as teams scored more goals, defencemen scored more goals relative to the forwards.

This does make sense; if your defensive players are focusing more on offence, your team will score more goals because you have more men trying to put the puck in the net. But of course, your opponents will also tend to score more goals, since you have less focus on preventing goals as well. For whatever reason the back players started to play offence more and more, resulting in more goals by their side and more goals by their opponents as well.

The following chart shows the year-by-year rate of team goals per game, divided by the average goals per game over the time period, and the proportion of goals scored by defencemen (which serves as a proxy for the defenders' relative involvement in the offence), divided by the average proportion over the time period. The goals per game ranges from a low of 3.03 in 1894 to a high of 7.73 in 1907. The percentage of goals scored by defencemen ranges from a low of 2.4% in 1892 to a high of 15.5% in 1908.

The relationship certainly isn't perfect, suggesting that there is more going on (not surprisingly), but it is quite a strong one. You can see this by looking at the chart; as goals per game increases, so does the percentage of goals scored by defencemen. The coefficient of correlation for these rates for this time period is 0.75, which is a very strong positive relationship.

So a big factor in why scoring increased so much over this time is that the point and cover-point, who previously focused on defence above all, became more and more involved in the offence. Whether this was a conscious decision on someone's part, or if it was a matter of the defensive players wanting to be more involved in the play, or something else entirely, we can't be sure. We know what was changing, but we don't know the true cause of the change.

The Scope of History

If you've ever discussed historical hockey players online, particularly in the context of comparing them to modern players, you may have come across the following comment, or something similar: "You can't really compare players from the past with those of today, because you didn't actually see the older players play." Similarly, when constructing some list of all-time something-or-others, someone will invariably say something like "I can only comment on the players I actually saw play."

In the context of hockey history, I find this type of comment baffling. History is not defined as "what you have witnessed personally." In studying the history of the game, of course you will encounter people and events that you have no direct personal experience with. To suggest that you can only make educated comments about things you have seen yourself is counter-historical. History involves the study of, well, history - things in the past, oftentimes beyond the memory of any living person, much less yourself. But events of the past do not suddenly become unknowable when the last direct witness passes - you just have to study the record in order to know it.

If you limit yourself to what you have seen yourself, I suppose you might feel more secure in the knowledge you do have (although the fact that you're relying on human memory has problems of its own). But let me tell you, if you choose not to study the history of the game because you can't "know" it the way you can with modern hockey, you're doing yourself a disservice. The history of the game is fascinating, and well-worth the effort of studying it.

Of course, if you're here reading this, you probably already agree.